**** BEGIN LOGGING AT Sat Dec 20 02:59:57 2008 Dec 20 04:49:54 Is there some documentation around for the gypsy dbus interface ? Dec 20 09:26:05 Hey Dec 20 09:26:13 hi Dec 20 09:26:27 quickdev: Have you managed to solve your problems with the muxer? Dec 20 09:26:47 PaulFertser_, no, I haven't. It seems to be not the muxer Dec 20 09:27:21 quickdev: Heh. So there's no new bugs there found? Dec 20 09:27:49 PaulFertser_, there's a bug somewhere, but I can't say where Dec 20 09:29:10 quickdev: The reason i ask is that when i saw a bug that was muxer related, i managed to solve it by analizing gsm0710muxd source, debug log and ogsmd debug log. Thought i could somehow help you. Dec 20 09:30:02 PaulFertser_, are you running shr? Dec 20 09:31:08 No, FSO HEAD for me. I'm afraid to use too many abstraction layers yet, before the basic functions are unstable. Dec 20 09:31:21 On a Debian system. Dec 20 09:34:48 quickdev: Moreover i don't want that much to setup OE yet ;) Dec 20 09:35:41 the problem was: if you have many contacts on your sim card, the result of request.getPhonebook('contacts') gets bigger than 4096. This is no problem if it's read in chunks, but sometimes a chunk is bigger than 4096 and gsm0710muxd only receives 4096 bytes, although there should be more. Thus frameworkd fails. Dec 20 09:37:16 quickdev: i list locking via aux after installing/starting scummvm which remaps the aux button, how do i get it back'? Dec 20 09:37:46 BluesLee, don't know Dec 20 09:38:11 quickdev: I can say that i haven't experienced this yet. I've 220 contacts on the SIM and i had the problem with the muxer. But after fixing it, it doesn't read more than 2047 bytes from the modem at a time. I'm unsure how can it happen that muxer sends more than that to the pty. I think i'll try to duplicate this. Dec 20 09:38:21 Ainulindale: do you know? Dec 20 09:39:11 PaulFertser_, is your fix upstream? Dec 20 09:39:54 quickdev: not sure about pyneo's version, but it's in 0.9.2.1 (IIRC) (so yes for FSO git). Dec 20 09:41:21 PaulFertser_, I'll write a little script later that has a while(1) { gsmsim.RequestPhonebook('contacts') }. That should reproduce it after some t ime Dec 20 09:43:47 I haven't tested raising GSM0710_BUFFER_SIZE yet. Only tried to use higher values for the other buffers which were 4096 bytes big Dec 20 09:43:55 afk for a while Dec 20 09:45:08 quickdev: what do you think should happen if the muxer constantly outputs 2047-bytes chunks to the pty and frameworkd is too slow to read it? Will the muxer block on writing? Then it'll lose the bytes from the modem? Else the bytes will be lost on writing to pty. Does it look like the case? Dec 20 09:56:24 quickdev: are there any plans to switch to newer kernel revisions? Dec 20 09:59:53 quickdev: And what is suspicious is that N_TTY_BUF_SIZE is exactly 4096 bytes. Dec 20 10:02:39 quickdev: I am afraid the constants you tried to raise are not effective at all. At least one of them (the buffer where you read() from the modem) is never effective, as the maximum transfer size is guarded by the free space in gsm serial buffer. Dec 20 10:03:43 quickdev: And the other constant can't be to blame also, as it is used to receive the data from the pty, not to send to it. Dec 20 10:13:59 quickdev: From the source it seems that ptys are opened in NONBLOCK mode (while serial is set to blocking, btw). I think the best you can do now is to add a verification to the write() in function extract_frames(), we should check how many bytes were actually written (not sure, might be fixed upstream at pyneo). Dec 20 10:15:34 quickdev: If my analisys is right, we can implement an additional large enough buffer to avoid overflowing the pty's. Dec 20 10:15:58 PaulFertser_, sounds fine Dec 20 10:18:13 PaulFertser_, could I ask you to make a patch? :) I'm far away from knowning gsm0710muxd's code Dec 20 10:19:30 quickdev: Ok, let's do as follows. I'll make a patch to add debugging output to that write() call and send it to you in some 20 min. Then we'll decide. Dec 20 10:20:44 quickdev: unfortunately the newest pyneo version seem to be the same in this regard. Dec 20 10:21:06 PaulFertser_, then we decide what? ;) Dec 20 10:24:46 quickdev: then we'll decide whether the reason i suspect is the case and what to do next ;) Dec 20 10:29:56 quickdev: So, please do as follows, apply http://pastebin.com/mc0969f6 , then enable gsm0710muxd debugging (start it with -vvv), make sure you have syslogd catching WARNING (and preferably DEBUG) messages from the muxer. Then if you see the message in the logs on failing attempt, it means we know what to fix. Dec 20 10:33:16 have to go for an hour, back later, sorr Dec 20 11:07:55 PaulFertser_, still here? Dec 20 11:08:06 quickdev: yes. Dec 20 11:09:48 PaulFertser_, I'm applying and testing it now Dec 20 11:11:21 quickdev: don't forget that the message that might be emitted is of WARNING level (so it'll probably go to /var/log/messages, not /var/log/debug). Dec 20 11:46:18 quickdev: It might happen that you won't be able to reproduce it with full debugging enabled. I think -vv is enough, it'll still allow logging of warnings while dropping DEBUG. Dec 20 11:46:50 PaulFertser_, yeah...reproducing it is quite hard now. It was easier before ;) Dec 20 11:47:55 quickdev: did you have full debugging turned on before? Dec 20 11:48:06 no, I had not Dec 20 11:50:27 quickdev: In fact i think you can run without -v's at all. Dec 20 11:51:30 quickdev: It's LOG_INFO by default and WARNING is more important, so it should get through. Dec 20 11:59:21 good morning doc Dec 20 11:59:36 or rather good evening Dec 20 11:59:48 PaulFertser_, http://rafb.net/p/BwQVvU87.html Dec 20 12:01:39 PaulFertser_, maybe we should look into errno? Dec 20 12:01:59 the buffer is probably full? Dec 20 12:02:45 quickdev: Yeah, we should have looked into errno but i bet it's EAGAIN. ;) Dec 20 12:03:05 we, would have bet, too ;) Dec 20 12:04:24 quickdev: I want to ask mickey|zzZZzz , does he also thinks we should buffer more at the muxer side? Dec 20 12:05:08 PaulFertser_, he told me to raise the muxer buffer, but you should ask him again Dec 20 12:05:52 quickdev: do you think profiling is doable directly on the FR, with output flowing through a USB console ? Dec 20 12:06:05 quickdev: The reason we lose bytes is not the muxer buffer, but the PTY buffer. He didn't know that, i guess. Dec 20 12:06:09 PaulFertser_: does andy-tracking poweroff on shutdown for you? Dec 20 12:06:40 PaulFertser_, yes, he didn't know that before Dec 20 12:06:49 mrmoku: nope. Dec 20 12:06:49 MarcOChapeau, profiling what? Dec 20 12:07:41 quite annoying to always have to pull the battery :( Dec 20 12:07:58 quickdev: well, 2 very different things... 1 would be general plain old C profiling with something like gprof. the other would be python apps (I'm thinking of frameworkd) Dec 20 12:08:05 mrmoku: do reboot instead of shutdown and hold AUX, poweroff from u-boot ;) Dec 20 12:08:21 MarcOChapeau, I don't now, try it then ;) Dec 20 12:08:55 PaulFertser_: hehe, does andy know about it? Dec 20 12:09:02 quickdev: I will :D Dec 20 12:09:06 mrmoku: and probably we should file a bug for kernel devs to fix. I'm not sure ;) Dec 20 12:10:04 PaulFertser_: compiling latest from now. If it still does not poweroff I will report it Dec 20 12:11:17 quickdev: i'll try to implement a proper buffering for writing to ptys probably later today. I think that the most important is that the bug was found, actually fixing it shouldn't be that hard ;) Dec 20 12:12:00 quickdev: BTW, does anyone know, what alternative muxer implementation Mickey is constantly hinting about? Dec 20 12:12:03 PaulFertser_, yeah, right. I'd appreciate it :) Dec 20 12:12:24 I don't know ;) Dec 20 12:15:57 I assume the in kernel one like on openezx (mux_cli) Dec 20 12:34:48 PaulFertser_, just to have it verified, it's EAGAIN Dec 20 12:35:02 quickdev: thanks :) Dec 20 12:37:28 PaulFertser_, is there a temporary fix? Is it possible to extend the PTY buffer a bit? Dec 20 12:38:03 anyone know how to get openmoko-icon-theme-standard2 to bitbake? Dec 20 12:38:05 http://pastebin.com/m7f7e94b7 Dec 20 12:38:50 quickdev: look, you don't want to loop on that write, because you can miss the data from the modem. And the only way to extend the PTY buffer is to recompile the kernel. Dec 20 12:42:47 thus we should extend gsmmuxd's buffer Dec 20 12:44:10 quickdev: and there's no buffer in the gsmmux for writing. We probably should implement them. Dec 20 12:48:15 PaulFertser_, that buffer should be allocated dynamically, right? Dec 20 12:48:28 because we can't set a limit Dec 20 12:49:32 quickdev: do you think in case frameworkd is not listening we should collect data for pty until the RAM fills Dec 20 12:50:30 no, but we should a least set a big limit of something like 32 KB Dec 20 12:50:44 maybe 16 is enough Dec 20 12:53:20 quickdev: for every pty, yes. I think we should alloc a buffer dynamically on allocating a new pty. Dec 20 12:53:40 PaulFertser_, is it much work to implement that write buffer? Dec 20 12:55:26 quickdev: not _that_ much but one needs to do it carefully, not to introduce any new bugs. Dec 20 12:56:12 yeah, please inform me as soon there's something to test ;) Dec 20 12:57:00 quickdev: sure. You should have told me earlier about your mux problems ;) Dec 20 12:58:54 PaulFertser_, I found the mux problems 2 days ago, because they only appeared in my new phone user interface and because they are not always appearing Dec 20 12:59:15 thus I wasn't able to report them earlier :) Dec 20 13:01:53 quickdev, PaulFertser_: that's a nice catch. what where the symptoms of those lost bytes ? Dec 20 13:03:21 MarcOChapeau, for example frameworkd wasn't able to do gsmsim.GetPhonebook('contacts') with many contacts on your sim under certain circumstances. Symptoms in SHR: segfaults ;) Dec 20 13:03:51 oh, yeah... really nice catch Dec 20 13:25:01 good morning mickeyl Dec 20 13:26:23 mickeyl, PaulFertser_ found out, that the TTY buffer is full under some circumstances. Thus write fails and the bytes are lost. Dec 20 13:27:12 morning, mickeyl :) Dec 20 13:27:26 morning folks Dec 20 13:27:29 good find Dec 20 13:27:32 strange though Dec 20 13:27:32 PaulFertser_, maybe explain it a bit more Dec 20 13:27:34 this is flowcontrolled Dec 20 13:27:36 isn't it? Dec 20 13:29:19 mickeyl: it's the PTY buffer that is used to transfer data from the muxer to the framework that gets overflowed. The reason of lossage is that gsm0710muxd just tries to write() into the pty, without checking how many bytes were written and not taking EAGAIN into account as well. Dec 20 13:30:26 hey dolf1074 Dec 20 13:30:31 hey Dec 20 13:30:50 http://alasal.be/openmoko/shr2/contacts/r2/ Dec 20 13:31:00 I took your comments into consideration Dec 20 13:31:25 PaulFertser_: ooh! Dec 20 13:31:29 ~lart gsm0710muxd Dec 20 13:31:30 * bzzbot does a little 'renice 20 -u gsm0710muxd' Dec 20 13:31:36 can you patch it? Dec 20 13:31:48 this will also be relevant for the alternative muxd i'm playing with Dec 20 13:32:23 mickeyl: you are writing a new muxer? Dec 20 13:32:41 dolf1074, maybe you misunderstood me. What I wanted to say: Shouldn't we place the A-Z bar horizontally at the top? And I think it's more consistent if we leave the NEW button at the bottom Dec 20 13:33:10 I'm stil considering the A-Z bar to the top Dec 20 13:33:20 bumbl: so to say Dec 20 13:33:21 I think it would eat to much space Dec 20 13:33:47 But I placed the new button to the top, so we have a details button when you select a contact Dec 20 13:34:07 bumbl: it's 100% transparent though, same dbus interface Dec 20 13:34:31 bumbl: and not much to do there, since we're reusing Nokia's code Dec 20 13:34:32 err Dec 20 13:34:35 Trolltech, that is Dec 20 13:34:37 :) Dec 20 13:34:46 dolf1074, when scrolling down, is the "Contacts New" disappearing? Dec 20 13:35:06 yes Dec 20 13:35:25 or shouldn't it? Dec 20 13:36:10 mickeyl: I think i can implement additional write buffering for the ptys. But do not wait for a working patch earlier than tomorrow's evening. Dec 20 13:36:26 PaulFertser_: take your time. great work Dec 20 13:36:39 * mickeyl doing ezx today anyways Dec 20 13:37:10 mickeyl: is qtopia's muxer really worthy? Don't you want to write a new one preferably from scratch? ;) Dec 20 13:37:47 i contrast to popular belief, I'm not suffering the not-invented-here syndrome :D Dec 20 13:37:53 * mickeyl loves reusing good code Dec 20 13:37:59 emphasis on _good_ Dec 20 13:38:18 quickdev: Did you already upload the newest code (with the fix of the segmentation fault) to unstable? Dec 20 13:38:19 That's why i ask about the quality. Dec 20 13:38:27 i'm just longing to have an alternative muxer implementation to make sure our muxer implementation is not the problem for some strange bugs Dec 20 13:40:29 mickeyl: I'm afraid i must admit that "our muxer implementation" was never meant to be used for long time. It's a temporary hack to get by while waiting for the "real thing" (kernel-space? heh;) ). Therefore i think that for such important core functionality we'd better have at least one decent muxer. Dec 20 13:40:46 dolf1074, I will comit a fix for #193 later. The segfault is because of the muxer problem mickeyl and PaulFertser_ just talked about. A fix will be there in 2-3 days :) Dec 20 13:41:04 ok Dec 20 13:41:30 quickdev: segfault? Are you really sure that buggy muxer should be able to segfault anything? Dec 20 13:41:36 It was just because I had another segfault and I cannot confirm it's the same as the others Dec 20 13:42:03 PaulFertser_: the kernel muxer is still on the agenda, and it looks good -- it will come. until then, I'm 100% behind gsm0710muxd, but I want to be able to crosstest w/ qtopiamuxd to pinpoint some "effects" to e.g. the modem-side's muxer implementation. Dec 20 13:43:37 PaulFertser_, frameworkds gets borked once there a incomplete messages. Thus invalid dbus responses are sent which should never appear. Anyway, you're right. It should not crash. Dec 20 13:44:14 dolf1074, imagine your scrolling down to see wether a contact is already there....and if it's not there, you have to scroll up again to see the new button. Dec 20 13:45:16 for the records, securing the lowlevel AT parser against invalid responses from the modem is a major undertaking, which I'd rather spend on not having the problem appear in the first place Dec 20 13:45:42 quickdev: true Dec 20 13:57:46 freesmartphone.org: 03mickey 07framework * ra06be62ed80d 10/framework/subsystems/ogsmd/ (3 files in 2 dirs): Dec 20 13:57:46 freesmartphone.org: ogsmd: move numberToPhonebookTuple from gsm.const into modems.abstract.modem Dec 20 13:57:46 freesmartphone.org: We need to be able to override it per-modem (Freescale Neptune *cough*) Dec 20 14:08:08 freesmartphone.org: 03mickey 07framework * rac11349cdcaa 10/framework/subsystems/ogsmd/modems/ (abstract/modem.py freescale_neptune/modem.py): Dec 20 14:08:08 freesmartphone.org: ogsmd: [FREESCALE NEPTUNE] override numberToPhonebookTuple Dec 20 14:08:08 freesmartphone.org: Storing international numbers works on this modem now. Dec 20 14:24:42 * bumbl sends mrmoku a festbock Dec 20 14:37:32 * mrmoku half drunken thanks bumbl :-) Dec 20 14:39:16 haha Dec 20 14:40:20 i immediatly thought of munich when i drank a "Maß" yesterday Dec 20 14:41:22 * mrmoku never understood why beer get's sold in 0.3 glasses ;) Dec 20 14:42:37 haha Dec 20 14:43:05 well for the ladies Dec 20 14:46:19 omg, I ate too much Cholent (http://en.wikipedia.org/wiki/Cholent ) I'm so stuffed that my tummy hurts. Dec 20 14:46:21 anyway never try to compete with one who is in a fraternity - they drink half an litre in 9s Dec 20 14:47:57 TAsn: it takes a long time to prepare :p Dec 20 14:48:09 yeah. Dec 20 14:48:17 And as it seems, a long time to digest too ;\ Dec 20 14:48:20 bumbl: yep lots of ladies I know drink more and faster than me ;) Dec 20 14:48:39 maybe that depends, because most of them are italian :p Dec 20 14:49:27 haha Dec 20 14:51:12 quickdev: I'm done with testing the Contacts app. http://spreadsheets.google.com/ccc?key=popRpaZG1txGaO8aIgaMNLg&hl=en Now going over on 'messages' part Dec 20 14:51:58 quickdev: btw. you showed me a screen on scap with a new dialer. why isn't that included in the new libframeworkd-phonegui-efl? Dec 20 14:52:24 quickdev: And I made new mockups: http://alasal.be/openmoko/shr2/contacts/r3/ Dec 20 14:52:39 dolf1074, which version are you currently using? Could you quickly do a screenshot of the dialer ui? Dec 20 14:54:32 I use the one in unstable. but how can I see the version? What opkg commands I have to do? Dec 20 14:56:00 dolf1074, unstable should have a dialer with white buttons, right? Dec 20 14:56:45 I has a dealer with black buttons and there doesn't come a red DEL button after you typed in something Dec 20 14:57:09 dolf1074, sorry, but then it's the old version Dec 20 14:57:29 strange Dec 20 14:57:32 I will look Dec 20 14:58:03 the problem could be, that in /etc/opkg/ shr-testing is configured. Maybe your package was downgraded or something like that to the shr-testing version. I don't know Dec 20 14:59:18 I had downloaded shr-testing and then adapted the /etc/opkg so it uses the unstable feeds and then I updated. But I will try to log into my freerunner and confirm Dec 20 14:59:25 dolf1074: could it be that you downgraded to shr-testing Dec 20 14:59:33 i have unstable and the new dialer Dec 20 14:59:45 s/have/had/ Dec 20 14:59:45 bumbl meant: i had unstable and the new dialer Dec 20 15:00:33 dolf1074: mockups look good - but I would prefer less colours Dec 20 15:04:20 okay I have still testing feeds in my opkg. (feel rather stupid) Dec 20 15:06:56 quickdev: But if I have the unstable feed and do opkg update/upgrade, I should have the newest version? Dec 20 15:07:21 try it Dec 20 15:07:40 it thinks nothing has to be upgraded Dec 20 15:12:32 bumbl: you don't like the color? or you don't like it having so much colours? Dec 20 15:18:55 freesmartphone.org: 03mickey 07framework * r1a4abcb52a44 10/framework/subsystems/ogsmd/ (3 files in 2 dirs): ogsmd: [ABSTRACT] strip " in response to +CPIN since some modems return this as a string Dec 20 15:18:56 freesmartphone.org: 03mickey 07framework * rc0d3cebcd788 10/ (4 files in 2 dirs): Dec 20 15:18:56 freesmartphone.org: ogsmd: [FREESCALE NEPTUNE] send the magic SIM initialization commands, Dec 20 15:18:57 freesmartphone.org: so that the GSM 07.07 commands for reading the SMS message book work ootb. Dec 20 15:18:59 freesmartphone.org: Override one channel's modemStateSimUnlocked() as well to send the mandatory Dec 20 15:19:01 freesmartphone.org: org.freesmartphone.GSM.SIM.ReadyStatus( True ) on time. Dec 20 15:32:10 freesmartphone.org: 03mickey 07framework * rebfe002459f0 10/docs/ogsmd/freescale_neptune/ (cfun.sample modem): document some more freescale neptune quirks Dec 20 15:35:02 freesmartphone.org: 03mickey 07framework * r15238037169d 10/docs/ogsmd/freescale_neptune/modem: freescale neptune: more quirks Dec 20 16:02:58 ptitjes, look at the git repo of pyphonelog, changed the 3 buttons to be exactly like the other apps (contacts/messages/etc) Dec 20 16:03:08 just like you wanted ;] Dec 20 16:03:40 oh, and changed today's date to show as 'Today, ' and not the full date. Dec 20 16:03:59 mwester: ping - i don't know if you are the right one to ask but: am I right that it would mean a tremendous amount of work to port the andorid java vm to x11? Dec 20 16:05:12 TAsn: scap it ... it*s hot Dec 20 16:06:35 bumbl, too tried now ;] Dec 20 16:06:47 heh Dec 20 16:06:51 I have to scap everything and upload to the wiki anyway, so I'll do it tomorrow or something. Dec 20 16:06:58 ah ok Dec 20 16:07:02 as I said earlier, the cholent made me feel BAD! Dec 20 16:07:26 and for some reason the accelerometers don't work anymore (testing...) Dec 20 16:09:24 phew, they work now. Dec 20 16:39:24 mickeyl, ping Dec 20 16:45:44 ptitjes, here? Dec 20 16:46:19 ptitjes, http://rafb.net/p/n4V6Zq16.html - phonelog seems to fail on the first boot of shr-unstable Dec 20 17:02:15 mwester, here? Dec 20 17:04:55 quickdev: are you sure you have current phonelog? ptitjes had an error closing the db in exit_cb (assuming this would be exiting ophonekitd) - but that is fixed... Dec 20 17:05:26 in b3961977c3c2b3ce8fabded3a7cbea9115b4ca6f Dec 20 17:05:49 ophonekitd - 0.0.1+gitr511+685f8399783a4cdeded0a462469112cb10d211b9-r12 - Ophonekitd daemon Dec 20 17:06:58 mrmoku, you are right. It's too old Dec 20 17:07:21 :-) Dec 20 17:33:06 mickeyl, could "c.remove( 'Z' )" in the channel.py affect things in a negative way? Dec 20 17:33:24 I mean is it a very ugly hack or can we leave it like that? Dec 20 17:33:43 quickdev: pong. Dec 20 17:35:24 mwester, I've got a gta02a04. ioctl seems to make the screen white. Is that the WSOD? Dec 20 17:35:37 when clicking on the screen it gets normal again Dec 20 17:35:47 it should be black / off instead of white Dec 20 17:36:26 i wouldn't say this is a wsod Dec 20 17:36:36 as it does not die Dec 20 17:36:39 I think that's a different problem, but it may be related to the kernel work done to try to fix the WSoD if it has just started doing this on recent kernels. Dec 20 17:40:10 * mrmoku thinks quickdev never suffered from *real* WSoD Dec 20 17:51:37 the white light at the end of the tube, hehe Dec 20 17:53:22 dcordes_: feel free to remove it in your modem abstraction Dec 20 17:53:28 i singleline it should stay Dec 20 17:53:31 s/i/in7 Dec 20 17:53:44 you will want to create an own modem anyways Dec 20 17:53:48 to support your custom commands Dec 20 17:55:48 mickeyl: I noticed two strange things today related to the absence of a GSM network. Dec 20 17:58:17 mickeyl: First: i suspended the phone in the morning and went to my job by the underground (where GSM coverage is spotty at best). And then i walked from the station to my workplace for about 20 minutes. After that i called my FR from the fixed phone and it was "out of coverage". Then i unsuspended it by pressing the power button, and in 2 minutes received an SMS that somebody (i.e. me) has called it. What do you think about it? Dec 20 17:59:46 mickeyl: And another one: on my way home i was unable to suspend it (exact reason is unknown yet). And at about the time i went underground i see that the +CMS ERROR: 321 Dec 20 18:00:24 answers in my log were gone and i had TIMEOUT 'AT+CMGL=2' ever since. Dec 20 18:01:28 can you send me the debug logs? Dec 20 18:04:36 mickeyl: ogsmd debug logs, yes. Btw, the reason why i was unable to suspend is TIMEOUT 'AT+CTZU=0' again. And i see that (though it was on another channel) i had an exchange AT+CMGL=2 -- +CMS ERROR:321 right after sending AT+CTZU=0 (and before the timeout). Dec 20 18:05:19 interesting. so after this it stays in the timeout mode forever? Dec 20 18:05:44 if the parser gets some bogus stuff, i could explain that Dec 20 18:09:02 mickeyl: Two different issues: First, timeout on AT+CTZU=0. It is a real timeout, the modem doesn't answer on the command. I'm afraid its AT command interpreter was confused :-/ Second, timeouts to AT+CMGL=2. This one starts at approx. the time i entered the underground (an hour later the unsuccessful suspend) and lasts till now (calling from the fixed phone gives "out of coverage again"). Dec 20 18:10:18 mickeyl: The frameworkd's parser doesn't seem to be confused in either case. Dec 20 18:11:07 mickeyl: Please tell me where you want my log to be sent, and whether you want the log to be "reasonably trimmed" and/or compressed. Dec 20 18:12:23 if you have private data in there, then send it to me directly (tar.bz2) Dec 20 18:12:29 if not, please attach it to a bug on trac.fso Dec 20 18:14:37 mickeyl: Do you have a particular bug in sight or should i create a new ticket? Dec 20 18:15:12 mickey|dinner: Bon appetit :) Dec 20 18:39:44 hello Dec 20 18:39:54 quickdev: did you succeed with the phonelog ? Dec 20 18:40:21 TAsn: that is cool! Dec 20 18:40:47 ptitjes, I'm glad you like it. Dec 20 18:40:54 IT TOOK ME HOURS TO ACHIEVE Dec 20 18:41:01 I experienced a stupid gtk behavior Dec 20 18:41:04 I couldn't solve Dec 20 18:41:19 until after hours of playing with it I succeeded. Dec 20 18:41:29 arf :) Dec 20 18:41:34 glad you succeed Dec 20 18:46:23 mickey|dinner, I sent a mail to the fso and moko ml Dec 20 18:47:58 PaulFertser: please create a new one Dec 20 18:48:00 dcordes_: k Dec 20 18:50:16 mickeyl: created. I hope you'll succeed in investigating these strange issues. Connection sharing, netdev led triggers etc are cool, no doubt. But basic functionality matters :( I very much suspect some more bugs in the calypso itself :-/ And lack of communication from TI really pisses me off. Dec 20 18:54:34 hmm Dec 20 18:54:40 apparantly there are subtle differences Dec 20 18:54:47 since some folks use the FR as their daily phone Dec 20 18:55:07 e.g. half of the FSO team Dec 20 18:55:51 mickeyl: They don't go by underground, sure ;) Dec 20 18:56:18 true :) Dec 20 18:56:24 no underground in Brunswick Dec 20 18:56:38 i promised to use the FR as a daily phone as well Dec 20 18:56:43 will start next year Dec 20 18:56:49 so I'm going to feel the pain Dec 20 18:57:42 Actually, i don't feel the pain. Permanent connectivity doesn't matter that much to me. But my logs are not faked :-/ Dec 20 18:58:08 ack Dec 20 18:58:46 hmm Dec 20 18:59:03 I think if you tell me exactly what tests with FR underground should i perform, i can do it. Dec 20 18:59:12 PaulFertser: what do you use for contacts and messages when you use it as a daily phone? Dec 20 18:59:23 bumbl: Zhone. Dec 20 18:59:33 hmm Dec 20 18:59:44 rwhitby: ping Dec 20 18:59:46 zhone is nice Dec 20 19:00:00 PaulFertser: thanks Dec 20 19:00:06 bumbl: my usage patterns are kind of humble. Receive 3 calls a day, make 1 at best. Probably receive a couple of sms's... Dec 20 19:00:07 but zhone still stores contacts, etc on the sim and not on flash Dec 20 19:00:10 we're probably going to write some monitorprogram Dec 20 19:00:16 that scans for fishy actions Dec 20 19:00:38 so I can't use it on a daily basis at the moment Dec 20 19:00:47 (just on stressless days) Dec 20 19:00:56 ptitjes, also wrote a workaround for the weight bug, this one is annoying. Dec 20 19:01:10 rwhitby: 140.211.169.169 is offline, anything i can do? Dec 20 19:01:21 mickeyl: wrt gsm0710muxd fix... Unfortunately, i can't code atm, too sleepy after a work day and a meal. But i think i'll manage to do it tomorrow. Dec 20 19:01:35 PaulFertser: ok, sounds good Dec 20 19:03:32 DocScrutinizer: whats that http://people.openmoko.org/joerg/SD-capacitor_GPS-fix/ ? Dec 20 19:04:25 gps hardware workaround Dec 20 19:04:35 a suugestion how you could possibly try to create a capacitor of around 10pF to stop the SD/GPS interference, without using an iron ;-) Dec 20 19:05:27 DocScrutinizer: there is a sw fix, are there any further advantages for me? Dec 20 19:05:34 It's not *that* serious - means it's never been tested or meant for an official recommendation Dec 20 19:06:01 The sw-fix is considered ok for everyday use. Dec 20 19:06:09 DocScrutinizer: i could handle that;-) Dec 20 19:06:10 anyone did an opkg upgrde on shr lately? should I go for it Dec 20 19:06:23 Zorkman: i did this morning Dec 20 19:06:27 BluesLee: so maybe give it a try Dec 20 19:06:54 DocScrutinizer: but if there is no advantage it makes no sense Dec 20 19:07:10 DocScrutinizer: if i would get a fixer fix maybe;-) Dec 20 19:07:11 there's very little advantage Dec 20 19:07:26 thats? Dec 20 19:07:34 mickeyl, could you comment on adding modems/msm7xxxx/ ? I would like to try it later. I'm not sure about the naming - msm7xxxx would be nice because it is used in OE for the kernel too. on the other hand the differences across the devices might become too big to keep it in one modem Dec 20 19:07:55 nope, it would just avoid loosing your fix while playback of mp3 or even video (continuous sdcard access) Dec 20 19:08:15 DocScrutinizer: but i also use scotch tape for the sim:-) Dec 20 19:08:39 DocScrutinizer: no problem as there is no real video joy on the freerunner:-) Dec 20 19:08:57 dcordes_: i'm going to call it qualcomm_msm for now Dec 20 19:09:02 what's a scotch tap Dec 20 19:09:04 e Dec 20 19:09:08 will create the base abstraction Dec 20 19:09:10 then you can continue Dec 20 19:09:14 the only scotch I know one can drink Dec 20 19:09:38 bumbl: klebeband Dec 20 19:09:39 dcordes_: btw., next time cc: smartphones-userland@linuxtogo.org as well Dec 20 19:09:43 bumbl: http://people.openmoko.org/joerg/sdcard-handle/20080801991.jpg Dec 20 19:09:50 (at leat i think that it what it is called in german) Dec 20 19:10:33 Zorkman: ah Dec 20 19:10:36 tesafilm Dec 20 19:10:50 DocScrutinizer: will there be other interesting differences in the A8? Dec 20 19:11:09 better mic Dec 20 19:11:11 hmm DocScrutinizer: I might try after xmas Dec 20 19:11:40 something to tiker Dec 20 19:11:43 +n Dec 20 19:12:57 dcordes_: what's the gprs node? Dec 20 19:13:01 smd1? Dec 20 19:13:20 DocScrutinizer: noticable difference between the mics? Dec 20 19:13:28 better == immune to MICBIAS ripple, a little more sensitive Dec 20 19:13:46 Zorkman: otherwise, no Dec 20 19:14:07 mickeyl, this name is too general imho. it's /dev/smd0 . the patchfile has everything I changed. (except for enabling debug and selecting singleline modem type in /etc/frameworkd.conf ) Dec 20 19:14:31 the gprs node is smd0 as well? Dec 20 19:15:17 no gprs data is /dev/smd1 . earlier in the ml discussion I posted the script we use to get online Dec 20 19:16:10 pppd /dev/smd1 debug defaultroute local usepeerdns name $2 noipdefaul Dec 20 19:16:18 http://headfuzz.co.uk/files/android/ppp.notsh Dec 20 19:18:32 mickeyl, I think the modem name should signify it's used on the msm7xxxx chipset devices Dec 20 19:19:06 mickeyl, brb Dec 20 19:19:09 i prefer qualcomm_msm Dec 20 19:19:19 we can change that when we see the first device != 7xxx Dec 20 19:20:40 i'd rather start broad and then specialize Dec 20 19:20:42 vs. otherwise Dec 20 19:25:22 freesmartphone.org: 03mickey 07framework * raf4a94a2016e 10/ (7 files in 3 dirs): ogsmd: [QUALCOMM MSM] new modem abstraction based on singleline Dec 20 19:28:56 mickey|tv: did I miss something? Dec 20 19:41:49 god: ping Dec 20 19:42:07 * Hire god|tv Dec 20 19:42:10 ops Dec 20 19:42:36 Hire: i am viewing dr.house, so stfu and go away Dec 20 19:43:06 god|tv: I seez, i will ping you after dr.house Dec 20 19:43:19 haha Dec 20 19:45:04 Hire: stop talking to yourself ;) Dec 20 19:46:39 indeed Dec 20 19:47:18 Hire: if you think you're god and have to watch tv... your problem - but don't talk to yourself :p Dec 20 19:47:27 you are right Dec 20 19:47:36 i am watching the SCIENCE on tv Dec 20 19:47:38 ( dr.house ) Dec 20 19:47:40 the EVIL Dec 20 19:47:49 talking to oneself is the first step of going round the bend Dec 20 19:47:57 i am a really a bad person Dec 20 19:48:18 and i want that ainu|fallback or ainulindale or quickdev will spanks my hardly Dec 20 19:48:19 * mrmoku is happy Hire is some hundreds of km south :P Dec 20 19:48:23 or Dec 20 19:48:25 for the sake Dec 20 19:48:36 i will can kill some persons Dec 20 19:48:39 for fun Dec 20 19:49:03 bumbl: I think it's too late... Hire is lost :-) Dec 20 19:49:09 someone says pestellince? it sounds great Dec 20 19:49:21 mrmoku: i am everywhere Dec 20 19:49:31 Hire: did you play too long with your FR today? Dec 20 19:49:36 no Dec 20 19:49:38 :( Dec 20 19:49:46 i used it only as nav thx to navit Dec 20 19:49:50 so why did you go crazy? Dec 20 19:49:56 ahh, and it ported you to hell? Dec 20 19:50:02 ainulindale finds funny to brick SHR Dec 20 19:50:14 i am the reaper Dec 20 19:50:30 i am ALWAYS crazy Dec 20 19:51:10 * mwester puts Hire on ignore, and cancels his plans for a trip to Italy next summer. Dec 20 19:51:17 :P Dec 20 19:51:21 cool Dec 20 19:51:38 mwester: not all italians are like Hire.... just about 95% ;) Dec 20 19:51:42 haha Dec 20 19:51:43 nice choice Dec 20 19:51:45 hehe! Dec 20 19:52:04 if you locate the remaining 5% you could have i nice trip :-) Dec 20 19:52:15 mrmoku: now i will drink a peroni for YOU Dec 20 19:52:19 Actually most of the Italians I met on the last trip seemed quite rational and non-crazy. Dec 20 19:52:23 prost Dec 20 19:52:29 cheers Dec 20 19:52:46 i am not rational Dec 20 19:52:49 mw|: where to italy have you been? Dec 20 19:52:53 (Except for the one that ran me off the road near Pizza) Dec 20 19:52:55 Hire: drink two.. somebody has to drink it... and I'm just happy it does not have to be me ;) Dec 20 19:52:55 ah sorry mw| Dec 20 19:52:57 i am stupid as ainulindale says Dec 20 19:53:06 mwester: where to italy have you been? Dec 20 19:53:06 mrmoku: indeed Dec 20 19:53:41 * mrmoku finished a nice Augustinger and is going downstairs to the fridge to refill his glass :-) Dec 20 19:53:45 -g Dec 20 19:53:46 bumbl: Milan to Pizza to Rome, and in that area (only 10 days; short trip) Dec 20 19:53:56 hmm Dec 20 19:54:35 Hire: how is weather in Rome? Dec 20 19:54:55 and by the way... if you see Berlusconi... SPANK HIM!!! Dec 20 19:54:55 i would agree with mrmoku but at least 100% of the Italians I met enjoy their life Dec 20 19:54:58 Went through Heathrow to get there, and again to get back --- so now I do not say that any American airports are the *ugliest* in the world; they simply cannot compare! ;) Dec 20 19:55:28 mrmoku: sadly, bad :( Dec 20 19:55:33 it is rain Dec 20 19:55:35 RAIN RAIN RAIN Dec 20 19:55:46 ( raining ) Dec 20 19:56:00 Berlusconi is my second cousin Dec 20 19:56:06 naaahhh, i am kidding Dec 20 19:56:07 raining here too (if it makes you feel better) Dec 20 19:56:10 mrmoku: have you heard the tactless and stupid joke about the Italians and Berlusconi? Dec 20 19:56:13 i will wait when the nature kills him Dec 20 19:56:31 mrmoku: no, because i am drink a stupid italian beer Dec 20 19:56:37 bumbl: I don't think that a joke about Berlusconi could even be tactless ;) Dec 20 19:56:39 drinking Dec 20 19:56:44 what the fu---- Dec 20 19:56:48 mrmoku: well Dec 20 19:57:07 Hire: nature won't kill him... will be your master for the next 20 years or so :( Dec 20 19:57:33 ahah Dec 20 19:57:38 it is a my problem :) Dec 20 19:57:41 you are safe Dec 20 19:57:42 :D Dec 20 19:57:45 bumbl: tell me I'm always interested in jokes about italians Dec 20 19:57:57 Hire: no I AM NOT... my wife is italian Dec 20 19:58:05 :D Dec 20 19:58:09 i know Dec 20 19:58:15 well it was something like that: why are the germans better than the italians? Dec 20 19:58:28 but now you and your wife live in Germany Dec 20 19:58:37 hmm, I know that is a fact... but why? Dec 20 19:58:45 Hire: yep Dec 20 19:58:50 and I have to fight with a stupid midget, Berlusconi :) Dec 20 19:59:13 political asylum Dec 20 19:59:14 because the germans only voted hitler once, the italians berlusconi twice Dec 20 19:59:22 quite tactless and stupid Dec 20 19:59:32 bumbl: uuhhh Dec 20 19:59:48 twice? Dec 20 19:59:51 are you sure? Dec 20 19:59:57 italians votes him some times :) Dec 20 20:00:06 hehe, would be nice only twice Dec 20 20:00:39 bumbl: FOUR times they did it Dec 20 20:00:45 oh Dec 20 20:00:46 then Dec 20 20:01:08 now you know how stupid... (Hire excluded of course - would never want to offend god) ;) Dec 20 20:01:28 the joke was not by me - a friend of mine told me yesterday Dec 20 20:02:01 you are right Dec 20 20:02:06 bumbl: read that http://de.wikipedia.org/wiki/Silvio_Berlusconi Dec 20 20:02:10 quite interesting Dec 20 20:03:39 wow half of the article is about critics Dec 20 20:04:16 actually they voted him only three times... but he had four governments Dec 20 20:05:33 * mrmoku is now going to the fridge to stop thinking about italian politics which makes him feel even more bad than the weather Dec 20 20:07:54 sicu, here? Dec 20 20:13:59 hmm well Dec 20 20:19:38 * Hire has prepared a cuba libre Dec 20 20:20:45 Hire: good taste Dec 20 20:20:52 have you tried cola rum? Dec 20 20:21:56 cuba libre is rum + rum :D Dec 20 20:22:02 rum + cola Dec 20 20:22:14 :p Dec 20 20:23:49 gimlet \o/ Dec 20 20:26:30 hmm well try vodka + vodke Dec 20 20:26:32 :p Dec 20 20:27:44 vodka + redbull Dec 20 20:27:53 so I will die young Dec 20 20:28:17 yep first the alcohol damages your liver and then the redbul damages your liver Dec 20 20:28:26 but first he will fly :-) Dec 20 20:28:47 you are right Dec 20 20:28:56 * Hire goes to open a pampero Dec 20 20:29:13 anybody interested in openmoocow? Dec 20 20:32:06 no :D Dec 20 20:32:38 well openmoocow is nice Dec 20 20:32:53 but I haven't yet found it useful :p Dec 20 20:32:57 I just built it on the buildhost... Dec 20 20:33:51 bumbl: no useful not... but my children like it :-) Dec 20 20:34:41 well as I don't have children Dec 20 20:54:55 mickey|tv: testing the modem in a few. I want to try the fso-image gui first Dec 20 20:57:39 me neither Dec 20 20:57:41 I am single Dec 20 20:58:20 brb Dec 20 21:02:19 Hey Dec 20 21:02:31 He's back! Dec 20 21:02:52 hey Dec 20 21:02:57 he is ALIVE Dec 20 21:03:07 who cares? :D Dec 20 21:03:08 :* Dec 20 21:03:10 Well I was here today Dec 20 21:03:14 I just didn'tspeak Dec 20 21:03:16 liar Dec 20 21:03:18 Had other stuff to do Dec 20 21:03:25 what? Dec 20 21:03:33 are you talking about REAL LIFE? Dec 20 21:03:40 that bugged app? Dec 20 21:03:46 closed-source too Dec 20 21:03:53 Hire: even if you don't... some people have a REAL LIFE ;) Dec 20 21:04:06 it is a lieeeeeeeeeeeeeeeeeee Dec 20 21:04:08 a LIEEEEEEEEEEEEEEEEE Dec 20 21:04:10 42 Dec 20 21:04:18 the real life doen't existtttttttttttttttt Dec 20 21:04:25 the earth is a big computer Dec 20 21:04:28 Well, I read, cooked, went to see my grand parents, that kind of thing Dec 20 21:04:29 and it uses windows :) Dec 20 21:04:33 for the ainulindale's sake Dec 20 21:04:36 I even cooked amaretti, Hire Dec 20 21:04:42 good :D Dec 20 21:04:55 some good bolognesi's amaretti Dec 20 21:05:05 and are them goods? Dec 20 21:06:22 Yes they were. Dec 20 21:06:36 Ainulindale: you could close #168 (pingus) Dec 20 21:06:43 was, oops Dec 20 21:07:23 well yes the world is a big blue screen of death Dec 20 21:07:38 mrmoku: how come you can't do that yourself, let's correct that instead Dec 20 21:08:01 Ainulindale: do you have used the lemon's peels? Dec 20 21:08:32 Ainulindale: I asked you to empower me... but you wanted me to get on my knees and I refused to ;) Dec 20 21:08:45 What? I asked for that? That doesn't sound like me Dec 20 21:08:51 bumbl: if you look up at a sunny day Dec 20 21:08:56 ~hail mrmoku Dec 20 21:08:58 (you should have the rights mrmoku) Dec 20 21:09:02 DocScrutinizer: :D Dec 20 21:09:05 you were full and drunk :) Dec 20 21:09:15 mrmoku: drunk was a figure of speech Dec 20 21:09:27 because with lemon's peels the amaretti becomes more sweet and fiable Dec 20 21:09:30 friable Dec 20 21:09:37 mrmoku: ainu doesn't ask, he expects ;-) Dec 20 21:09:56 Hire: well I'll try that but I don't put lemon in mine Dec 20 21:10:04 mrmoku: why do you refuse? Dec 20 21:10:08 are you talking drinks again or still? Dec 20 21:10:09 ah ok ;) Dec 20 21:10:22 no wonder the openmoko gets nowhere ;) Dec 20 21:10:31 bumbl: because my knees hurt - I'm old :p Dec 20 21:10:38 How old? Dec 20 21:10:39 Fossi1: again and still Dec 20 21:10:43 Ainulindale: 39 Dec 20 21:10:49 mmm damn Dec 20 21:10:52 ah so wurp2-old Dec 20 21:10:54 not mwester-old Dec 20 21:10:56 i always forget that i'm on vacation Dec 20 21:11:06 how old is mwester-old? Dec 20 21:11:09 i could actually check out and build shr Dec 20 21:11:25 nobody is gonna call me in a while Dec 20 21:11:45 mrmoku: I would usually say 42 Dec 20 21:11:48 But in this case I don't know Dec 20 21:11:53 hehe Dec 20 21:12:09 well 42 is not the answer to all questions Dec 20 21:12:12 three years to go and I will be the answer to everything :-) Dec 20 21:12:18 sure it is Dec 20 21:12:50 usual, you only need a fairly broad context Dec 20 21:12:54 *usually Dec 20 21:12:58 (damn drinks) Dec 20 21:13:49 Hey did someone regenerate an image? Dec 20 21:14:02 no Dec 20 21:14:05 spank quickdev Dec 20 21:14:11 Well, no, that's good like that Dec 20 21:14:17 Regenerating an image now has no meaning Dec 20 21:14:19 lol - kids Dec 20 21:14:29 DocScrutinizer: kids? Dec 20 21:14:30 why not? Dec 20 21:14:38 it is always a good idea to spank him Dec 20 21:14:57 Indeed, indeed Dec 20 21:15:36 Do someone know if the gsmmuxer bug was fixed? Dec 20 21:16:15 Ainulindale: I think the root of evil is found (pty buffer overflow) but not yet fixed Dec 20 21:16:23 Ok good Dec 20 21:16:51 PaulFertser wanted to fix it for tomorrow evening IIRC Dec 20 21:17:39 Ainulindale: why there is no point in regenerating the image? Dec 20 21:18:19 because we're still in testing for quickdev's UI Dec 20 21:18:25 and I'd like to wait for the gsmmuxer patch Dec 20 21:19:32 furester? Dec 20 21:19:38 it is away Dec 20 21:20:06 Ainulindale: no PaulFertser Dec 20 21:20:20 are they still not in testing Dec 20 21:20:23 well Dec 20 21:20:30 mrmoku: I was asking for furester, had a question for him :-) Dec 20 21:20:38 ah, hehe Dec 20 21:20:40 (about latest enotes revision) Dec 20 21:20:47 bumbl: news in omoney? Dec 20 21:20:53 While I'm at it Dec 20 21:20:58 no Dec 20 21:21:01 or yes Dec 20 21:21:13 waiting for a new python-elementary revbump Dec 20 21:21:35 well Dec 20 21:21:42 it was promised for yesterday Dec 20 21:21:47 mrmoku: eager to help me with that? Dec 20 21:21:54 bumbl: well Dec 20 21:22:01 this implies EFL_SRCREV revbump IIRC Dec 20 21:22:11 i will replace etk with elementary Dec 20 21:22:27 Ainulindale: arghh, did that today... current EFL is bad Dec 20 21:22:33 Ainulindale: python-elementary is hosted somewhere different Dec 20 21:22:42 mrmoku: that's why we had to ping raster Dec 20 21:22:49 this would solve the locale problem too Dec 20 21:22:59 bumbl: but it has dependencies Dec 20 21:23:34 code cleanup is going on - i will push the changes maybe tomorrow Dec 20 21:23:35 Ainulindale: there is some libefreet changes that seem to make the icons in illume launcher disappear Dec 20 21:24:15 Ainulindale: well the current rev does not work with the current efl rev anymore so ... Dec 20 21:24:29 Ainulindale: how can I help, btw.? Dec 20 21:24:44 mrmoku: well, testing on unstable Dec 20 21:25:00 * mrmoku is running only unstable :-) Dec 20 21:25:25 I just flashed todays unstable image... Dec 20 21:25:25 Hire: is it possible to get a xcf file of one of your icons? Dec 20 21:25:44 Ainulindale: does not work good though Dec 20 21:25:59 mrmoku: well unstable EFL_SRCREV is not redefined Dec 20 21:26:08 ophonekitd is crashing with dbus quark now and then Dec 20 21:26:13 Huh? Dec 20 21:26:15 * mwester checks the calendar to see how old he is. Dec 20 21:26:16 bumbl: sure Dec 20 21:26:29 Ainulindale: yeah, but I forgot to include the normal sanerev one :p Dec 20 21:26:48 heh Dec 20 21:26:59 before I had a homebuilt unstable image that was working great though... Dec 20 21:27:08 bumbl: hiretto.altervista.org/base.tar Dec 20 21:27:16 fixed the Makefile to include both for unstable Dec 20 21:27:19 feel free to use it Dec 20 21:27:24 brb Dec 20 21:27:32 Hire: gracie Dec 20 21:27:50 -c +z Dec 20 21:27:57 * mwester passed Life, the Universe, and Everything 5 years ago Dec 20 21:28:17 mwester: now on to the next level ;) Dec 20 21:28:29 mwester: next step is "Pastis" (51) Dec 20 21:28:32 (french alcohol) Dec 20 21:28:58 grazie Dec 20 21:30:33 hmm base.tar is empty Dec 20 21:31:15 hmm now it works Dec 20 21:31:17 strange Dec 20 21:31:40 Ainulindale: one difference is, I was using frameworkd-devel Dec 20 21:32:45 bumbl: remember, Hire is EVIL. So better double-check the content :p Dec 20 21:33:39 hmm i must get an antivirus Dec 20 21:36:16 indeed :) Dec 20 21:36:20 i am a bad person Dec 20 21:36:41 michael jackson said the same Dec 20 21:38:12 and michael jackson turned out to be *very* evil. Dec 20 21:38:49 i dont think that hire will sell his nose to the devil;-) Dec 20 21:38:57 but he was very good to all the litte children *scnr* Dec 20 21:39:23 yes Dec 20 21:39:32 but i don't want to do sex with children Dec 20 21:39:52 i want to do sex only with women, french chicks and ainulindale Dec 20 21:39:59 :-) Dec 20 21:40:03 and irish women Dec 20 21:40:19 well i would spare Ainulindale Dec 20 21:44:52 i am tired Dec 20 21:45:12 go to bed Dec 20 21:45:32 5-6 hours sleep daily Dec 20 21:46:03 three days long Dec 20 21:46:34 cu tomorrow Dec 20 21:47:15 mrmoku: if you're interested in doing some C I may have something you could be interested in doing Dec 20 21:48:20 I also need someone who'd be inclined to do some python for frameworkd oriented stuff =) Dec 20 21:51:12 * mrmoku prefers C a lot :-) Dec 20 21:51:28 well we should write a "I want you for SHR army" Dec 20 21:51:55 Ainulindale: what would be the C stuff? Dec 20 21:53:52 well Dec 20 21:53:56 Still wondering about this one Dec 20 21:54:08 But I'd like to do a band aid to support names in ophonekitd Dec 20 21:54:19 local contact cache, and then name to number functionnality Dec 20 21:55:08 makes sense to do it in ophonekitd and not the frontends Dec 20 21:55:22 Indeed but makes no sense with opimd Dec 20 21:55:31 So if it's quick and fun for you Dec 20 21:55:44 I'll suggest you do it if you're interested in that :-) Dec 20 21:58:35 ladies Dec 20 21:58:46 I will get some real life this evening Dec 20 21:58:52 with the special pampero evening Dec 20 21:59:04 so... see you, have a nice evening Dec 20 22:00:01 Hire: have fun Dec 20 22:00:16 yep hf Dec 20 22:00:19 sure i will Dec 20 22:00:22 mickey|tv, errors on starting frameworkd. see ml Dec 20 22:00:25 ( i think ) Dec 20 22:04:18 already replied Dec 20 22:04:37 you mixed frameworkd with frameworkd-devel Dec 20 22:04:38 that's what i would guess from the error Dec 20 22:05:39 Ainulindale: that cache would be used in ophonekitd_call_status_handler to pass on the name instead of the number? Dec 20 22:06:11 yes Dec 20 22:06:20 and in the end a structure with both Dec 20 22:06:35 I think this would be wise to prepare stuff for opimd Dec 20 22:06:48 hey mickey|tv o/ Dec 20 22:08:02 mickey|tv, it has subsystems/ogsmd/modems/qualcomm_msm Dec 20 22:08:28 could be Dec 20 22:08:36 the controller might be different Dec 20 22:08:41 or the starting binary Dec 20 22:08:45 err... script Dec 20 22:09:53 let me check what I did in Oe Dec 20 22:11:57 mickey|tv, I installed frameworkd_0.8.4.4+gitr0+8d5b994d29a97b47c00f041520b76be64bdfe865-r0.1_armv6-novfp.ipk in the image and updated the framework folder manually from my local copy of the framework main git Dec 20 22:11:58 Ainulindale: would it be preloaded or dynamically filled when needed Dec 20 22:12:40 dynamically me thinks Dec 20 22:12:50 Well, that's your choice but dynamic is bad Dec 20 22:12:53 right.. in my last image I did the same and had to update to frameworkd-devel in the image Dec 20 22:12:54 (IMHO) Dec 20 22:12:57 Because it takes time Dec 20 22:13:04 And in between you don't have a window to do stuff Dec 20 22:13:24 I'd rather wait 3 seconds at startup than 1 second each call Dec 20 22:13:39 dcordes: yep, that's not enough Dec 20 22:13:44 you need to update all files Dec 20 22:14:06 yeah, one second on each call would be too much... right Dec 20 22:14:07 at least include frameworkd itself Dec 20 22:14:20 I will build the ipk and scp it over Dec 20 22:15:58 mrmoku: fire your thoughts here Dec 20 22:16:04 I may be wrong Dec 20 22:16:08 Even though on this one I think I'm not =) Dec 20 22:16:17 What I'm looking for, anyway Dec 20 22:16:22 I just think preloading it does not scale... Dec 20 22:16:31 Is a structure we will pass to phonegui Dec 20 22:16:42 Which will be usable once opimd hits the image Dec 20 22:16:50 Why wouldn't it scale? Dec 20 22:17:23 which will be hopefully soon Dec 20 22:17:31 with sim contacts it's fine... Dec 20 22:17:40 if it will be obsoleted with opimd fine also Dec 20 22:18:11 mrmoku: the structure itself won't be obsoleted Dec 20 22:18:15 the preloading will, yes Dec 20 22:18:20 ok, then Dec 20 22:18:22 that's why I told you, if you think it's fun and quick Dec 20 22:18:23 do it Dec 20 22:18:31 if you think it'll take you a huge amount of time and it'll be boring Dec 20 22:18:36 just find something more useful :-) Dec 20 22:18:52 And just so you know Dec 20 22:18:58 I don't do it myself because I find that boring =) Dec 20 22:19:16 hehe Dec 20 22:20:40 mickey|tv, is it necessesary to install frameworkd-devel-config ? Dec 20 22:22:44 yes Dec 20 22:22:49 it contains valid configuration and rules Dec 20 22:22:57 NoProvider: frameworkd-devel-config Dec 20 22:25:45 Hi all, I have some n00b questions. I'm trying to re-build wxwidgets with Unicode support. Seems vairly simple from the wxwidgets side, just adding the --enable-Unicode flag to compile options. my problem is setting up the build environment. I've tried both make image and make qemu and both break in some way. Any ideas? Dec 20 22:33:33 mickey|tv, http://rafb.net/p/pduEXO23.html Dec 20 22:39:01 grr, packaging error Dec 20 22:39:08 please grab that file from git Dec 20 22:39:14 i'll fix the recipe in the meantime Dec 20 22:43:07 ok Dec 20 22:58:57 Ainulindale: RetrievePhonebook is not (yet) implemented in libframeworkd-glib? Dec 20 23:02:04 hello, The passwords are blank on all distribs in order to ssh into the device...how does blank password line looks like in /etc/shadow ? Dec 20 23:02:13 hmmm, looks like I need to add an item to CFLAGS, any idea how to do that? I've tried build/conf/local.conf and setup-env Dec 20 23:08:33 try Dec 20 23:08:46 just pulled Dec 20 23:08:48 EXTRA_OECONF = "--enable-unicode" Dec 20 23:10:25 mickey|tv, what's the correct lcoal.conf line to make frameworkd build frameworkd-devel again? Dec 20 23:11:00 PREFERRED_PROVIDER_frameworkd = "frameworkd-devel" Dec 20 23:11:08 unfortunately I'm not up to building wxwidgets yet, make image still fails compiling GCC because it doesn't have -D_FORTIFY_SOURCE=0 Dec 20 23:11:33 i remember there was a thread about that on oe-devel, might be worth searching the lis Dec 20 23:11:34 t Dec 20 23:11:57 * mwester wonders why somebody doesn't just contribute some patches to fix those broken open() calls Dec 20 23:12:52 mickey|tv, I build the fso-image-light with the following lines and wonder why it gave me frameworkd Dec 20 23:12:55 PREFERRED_PROVIDER_frameworkd = "frameworkd-devel" Dec 20 23:12:55 require conf/distro/include/fso-autorev.inc Dec 20 23:13:04 mrmoku: Err yes it is Dec 20 23:13:16 mwester: well there is the say: if you want that it is done right do it yourself Dec 20 23:13:20 mrmoku: else how would we be able to have contacts? Dec 20 23:14:10 Ainulindale: retrieve_entry I do find... but not the whole phonebook... how is it called? Dec 20 23:15:07 mickey|tv: is it likely that qualcomm_modem will work on the cdma versions also? Dec 20 23:15:17 dcordes: no idea, it works for me here :/ Dec 20 23:15:22 mrmoku: errr Dec 20 23:15:26 no idea =) Dec 20 23:15:31 :D Dec 20 23:15:46 dcordes: you can change DEFAULT_PREFERENCE in the file as well; worth a try Dec 20 23:16:27 mrmoku: though quickdev used it in libframeworkd-phonegui-efl Dec 20 23:16:30 tmzt_: unlikely, I'm not even sure if CDMA versions use an AT interface Dec 20 23:16:47 [for communicating with the TA] Dec 20 23:16:55 tmzt_, of course it works Dec 20 23:17:06 mrmoku: if we didn't follow API change on this part feel free to patch it :-) Dec 20 23:17:20 tmzt_, I expect it to work on a vogue the same way it works on my kaiser Dec 20 23:17:35 amazing -- i stand corrected. looking forward to it then Dec 20 23:17:50 does this hold for all cdma modems? Dec 20 23:18:20 mickey|tv, iirc the dialing pat on msm7500 is the same as on 7200 Dec 20 23:18:29 cool Dec 20 23:19:36 I can ask somebody with a vogue or titan to try Dec 20 23:24:37 tmzt_, dzo's ril allows the kaiser and vogue to make calls.. Dec 20 23:25:23 tmzt_, in future I would like to enable the cdma specific commands in the qualcomm_msm fso radio too Dec 20 23:25:40 I know e.g. sms is different Dec 20 23:26:45 mickey|tv, They use the AT-command set Dec 20 23:28:01 Ainulindale: ok, though i'm off for today - will continue tomorrow Dec 20 23:29:04 right, dcordes: does the audio work with proc_comm? Dec 20 23:29:18 I wonder if these are too big to stick into the GTAv03: http://www.sierrawireless.com/product/fwt.aspx Dec 20 23:30:35 tmzt_, I don't know how to set audio routing on the kaiser. dzo told me trout uses rpc. when I just dial, it uses the speaker Dec 20 23:30:46 i.e. setting no state. so speaker mode seems standard Dec 20 23:30:47 alright, yeah, they are, they're huge. Dec 20 23:31:33 Sargun, those have qualcomm too Dec 20 23:31:50 dcordes, ...What's unique about that? Dec 20 23:32:08 Well, not unique? Dec 20 23:32:18 ok, how does the ril do it? can we make that work in python? Dec 20 23:32:19 it just came to mind Dec 20 23:32:46 dcordes, Aren't almost all EV-DO devices qualcomm, I didn't think Qualcomm sold much IP Dec 20 23:33:04 tmzt_, I'm sorry, I confused dzo and swetland... Dec 20 23:33:17 oh, right Dec 20 23:33:45 Sargun, I don't know Dec 20 23:34:01 Well, does the SoC do miniPCI? Dec 20 23:34:27 dcordes: does kaiser have /sys/class/vogue_hw/* ? Dec 20 23:34:54 I use the old htc-msm kernel now. it doesn't build vogue-hw.c Dec 20 23:35:12 because of mmc problems? Dec 20 23:35:16 51mm x 30mm x 4.5mm -- Too big? Dec 20 23:36:04 tmzt_, I don't know why Dec 20 23:36:22 I mean you use it because of that? Dec 20 23:36:35 Sargun: is it cardbus? Dec 20 23:36:55 tmzt_, no Dec 20 23:36:59 tmzt_, Mini PCI Express Dec 20 23:37:10 does it use the usb pins? Dec 20 23:37:12 tmzt_, yes. with sdhc it works very stable Dec 20 23:37:35 as long as you don't read and write a lot at the same time Dec 20 23:38:37 dcordes: with vogue_hw it appears you write 2 to audio to start the call 5 to end it Dec 20 23:39:14 mickey|tv, I updated frameworkd-devel with frameworkd-devel_0.8.4.9+gitr2+af4a94a2016e3e5a7fcf370932980c1d5ca2dac1-r0.1_armv6-novfp.ipk and it still lacks the file Dec 20 23:39:56 tmzt_, in htc-vogue branch? Dec 20 23:39:56 the file is in frameworkd-devel-config Dec 20 23:40:03 ok installing Dec 20 23:40:16 tmzt_, yeah Dec 20 23:40:39 it's a msm6550? Dec 20 23:41:38 My wishlist for a GTA 03: Virtex-3 FPGA at a 100 mhz clock (At minimum), fusion reactor, 800x600 4.3 inch screen. Dec 20 23:41:38 mickey|tv, still getting NoProvider: frameworkd-devel-config Dec 20 23:41:39 :-P Dec 20 23:42:05 yay, GCC built successfully Dec 20 23:42:15 no idea bout that Dec 20 23:42:40 perhaps your ARCH has a mismatch Dec 20 23:42:49 the config packages have another packaging ardh Dec 20 23:43:17 must be something like that. local oe git is clean Dec 20 23:43:25 and uptodate Dec 20 23:43:36 err, wait Dec 20 23:43:40 you try bitbake frameworkd-devel-config? Dec 20 23:44:12 yea :) Dec 20 23:44:14 ah. that's wrong. frameworkd-devel-config is a subpackage, not a provider. it's emitted by frameworkd-devel. Dec 20 23:44:29 bitbake providers can emit multiple packages. Dec 20 23:44:39 (and they usually do) Dec 20 23:45:23 k Dec 20 23:54:40 tmzt_, I think these states are needed to switch between phone audio use and normal audio use Dec 20 23:55:39 right, but you need them in the python code to make a call I thought Dec 20 23:56:23 no arm9 takes total posession and uses HANDSFREE automagically Dec 20 23:56:31 n8 Dec 20 23:56:54 it closes and opens the audio routing. I just dial. Dec 20 23:57:10 mickey|zzZZzz, n8 Dec 20 23:58:17 tmzt_, I'm wondering why it defaults to HANDSFREE Dec 20 23:58:45 bluetooth? Dec 20 23:59:01 maybe it just initializes the codec like that Dec 20 23:59:11 can you get vogue-hw.c to build on htc-msm? Dec 20 23:59:35 I can try tomorrow Dec 21 00:00:35 it's all proc_comm no? Dec 21 00:06:26 yes Dec 21 00:26:40 hmm, now qemu-native_0.9.1 breaks Dec 21 00:27:45 qemu-0.9.1/linux-user/syscall.c:4814: error: dereferencing pointer to incomplete type et al Dec 21 00:33:29 mickey|zzZZzz, the frameworkd -s ogsmd output looks ok but cli-framework can't connect to the bus. don't know what's wrong. will retry tomorrow Dec 21 00:37:43 urz: put this in your local.conf file: Dec 21 00:37:47 ENABLE_BINARY_LOCALE_GENERATION = "0" Dec 21 00:39:02 k, retrying Dec 21 00:54:34 bbiab Dec 21 01:17:30 Hey **** ENDING LOGGING AT Sun Dec 21 02:59:57 2008