**** BEGIN LOGGING AT Tue Feb 26 10:59:56 2008 Feb 26 11:44:44 Hi... I've flashed the 2008-02-20 image Feb 26 11:45:20 it seems to work just fine, but it doesn't get the PIN correctly... it asks... I give it, but it acts as if the PIN was incorrect Feb 26 11:45:34 I've already tried the sim in other phone Feb 26 11:45:39 and the pin is correct Feb 26 11:45:59 any idea why is this happening? any workaround? Feb 26 11:46:32 what if you disable the pin Feb 26 11:46:40 can I/ Feb 26 11:46:41 ? Feb 26 11:46:59 I mean... is it possible to use a gsm sim without a pin? Feb 26 11:47:07 yes Feb 26 11:47:07 i think someone said you had to enter pin manualy Feb 26 11:47:13 Mine has no pin Feb 26 11:47:46 13:43:26 < teprrr> I get the insert pin dialog, but it doesn't accept my pin code Feb 26 11:47:49 14:10:32 < teprrr> ahh. got it. had to enter the pin manually with libgsmd-tool :P Feb 26 11:48:24 SpeedEvil, but considering mine has... is it possible to remove pin checking from a working sim card? Feb 26 11:48:31 zash, ok... I'll try that... Feb 26 11:48:51 ruoso: i can disable pin in my 7 years old phone Feb 26 11:49:25 I have no idea. Feb 26 11:51:44 zash, hmmm... I don't think my phone has a way to do that... could you ellaborate on that? Feb 26 11:53:19 zash, hmm... I've just found it... Feb 26 11:53:21 thanks Feb 26 11:53:23 :D Feb 26 11:56:53 ruoso, someone said a few days back you had to enter the pin manually into gsmd or whatever Feb 26 11:57:14 someone had similar problems, i think there is a irc log for this channel somewhere Feb 26 11:57:23 12:47:46 < zash> 14:10:32 < teprrr> ahh. got it. had to enter the pin manually with libgsmd-tool :P Feb 26 11:57:33 ;) Feb 26 11:57:34 oh, right Feb 26 11:57:38 sorry :) Feb 26 11:57:41 i just woke up Feb 26 11:57:54 good morning too you then Feb 26 11:58:05 thank you! it is a good morning Feb 26 12:01:36 Blastur, I've disabled the pin in the sim card... but thanks anyway Feb 26 12:07:33 openmoko: 03werner * r4118 10/trunk/src/target/u-boot/patches/ (gta02-splash.patch uboot-gta02.patch): (log message trimmed) Feb 26 12:07:33 openmoko: Driving nDL_GSM low caused the GSM modem to inject communication requests Feb 26 12:07:33 openmoko: into the audio system, causing the characteristical clicking sound of a Feb 26 12:07:33 openmoko: GTA02 starting up, and these requests also caused lots of interrupts on Feb 26 12:07:35 openmoko: JACK_INSERT, slowing down the CPU. Feb 26 12:07:37 openmoko: This patch sets nDL_GSM high by default, and lowers it only when we turn Feb 26 12:07:39 openmoko: the modem on or off. (Under Linux, there nDL can be controlled separately Feb 26 13:42:02 hi to all Feb 26 13:43:02 i want to know if there is some project of interest Feb 26 13:43:08 that i can play for openmoko Feb 26 13:43:54 !!!! Feb 26 13:45:13 for example a porting of vlc Feb 26 13:45:16 can interest? Feb 26 13:46:31 why nobody speak? Feb 26 13:46:54 check projects.openmoko.org Feb 26 14:06:27 hi Feb 26 14:06:32 mickeyl: Ping? Feb 26 14:06:52 mickeyl: Are you coming to Austin for the LF Collaboration Summit? Feb 26 14:07:04 I'd be interested in having you there for the GNOME Mobile meeting Feb 26 14:09:28 dneary: hi. sorry, but i already booked a vacation during this time frame. besides, i promised my wife and bosses to travel less this year and do more coding ;) Feb 26 14:09:48 i did not yet decide on guadec though Feb 26 14:10:06 will there be a GMM as well there? Feb 26 14:11:02 mickeyl: go to guadec Feb 26 14:11:14 hrw: will you be there? Feb 26 14:11:18 mickeyl: it will be probably the only one conference for me in this year Feb 26 14:11:21 ah Feb 26 14:11:23 wait... you don;t code on those long distance flights?! Feb 26 14:11:50 mickeyl: guadec+oedem probably Feb 26 14:12:10 ScaredyCat: i don't. and i really hate flying more than 6 hours Feb 26 14:12:19 unfortunately FRA -> TPE is 14 Feb 26 14:12:32 lucky that you have direct flight Feb 26 14:12:36 right Feb 26 14:12:39 could be worse Feb 26 14:12:41 but you have those foldy out beds ? Feb 26 14:12:51 for me it would be POZ -> FRA -> TPE probably ;D Feb 26 14:12:53 ScaredyCat: heh. this is economy class Feb 26 14:12:58 or do they make you travel cattle Feb 26 14:13:00 :( Feb 26 14:13:04 of course :/ Feb 26 14:13:09 OM Inc. is hardly profitable yet Feb 26 14:13:56 aye I guess so... Feb 26 14:13:56 mickeyl: cna you reveal total sales of GTA01? Feb 26 14:14:50 SpeedEvil: i'm afraid not. i can only say we're sold out Feb 26 14:14:56 are you going to computex mickeyl? Feb 26 14:15:45 ScaredyCat: no. unless it's not super-super-mega-important, i will never visit taiwan in summer again. the climate is reducing my work efficiency a lot Feb 26 14:15:56 thermal throttling Feb 26 14:16:00 :) Feb 26 14:16:01 and the AC does not help a european... Feb 26 14:16:16 will FIC be there do you know? Feb 26 14:16:23 sure Feb 26 14:16:38 good stuff Feb 26 14:29:40 Marcel_M: Thanks!!! Feb 26 14:35:04 mickeyl: OK - try to make GUADEC then :) Feb 26 15:18:14 hey all, gsoc has been announced. Tell me OM will be participate to get in some cool code :D Feb 26 15:21:08 they did participate last year... Feb 26 15:21:56 yeah Feb 26 16:03:59 SpeedEvil: OSM-related company in London, http://www.zxv.ltd.uk/blog/ Feb 26 16:04:08 SpeedEvil: they're currently recruiting for J2ME devs Feb 26 16:13:36 Stephmw: Any idea of what the pay is like? Feb 26 16:14:05 I'm a java developer currently infatuated with the Neo who wants to move to London, so it sounds pretty exciting to me :-) Feb 26 16:14:49 On a totally different topic, is it accurate to say that the worst that could happen from my phone going into suspend is that I might lose a call? Feb 26 16:15:05 Resume works consistently, and leaves the gsm in a good state to receive/make new calls? Feb 26 16:15:14 wurp2: you'll find a couple of ZXVers on #osm on oftc Feb 26 16:18:00 cjb_ie: cool, thanks Feb 26 16:22:18 wurp2: good lord, I'd have no idea Feb 26 16:22:36 is Openmoko doing anything for GSoC? Feb 26 16:22:40 wurp2: I imagine salaries with the required experience might start around £24k Feb 26 16:23:03 wurp2: how's your J2ME? Feb 26 16:30:56 Stephmw: way too good to work for 24k pounds ;-) Feb 26 16:31:23 actually, I haven't focused on j2me; I'm primarily a server side guy Feb 26 16:31:39 wurp2: http://www.zxv.ltd.uk/jobs/java-developers-needed/ is the posting Feb 26 16:31:51 So I imagine both of us would be dissatisfied, at least superficially Feb 26 16:31:54 wurp2: if feels like a junior position, but who knows, maybe there's something better? Feb 26 16:31:55 cool, thanks! Feb 26 16:32:30 Yeah, that looks junior, but I think I'll send 'em a resume anyway Feb 26 16:32:49 Oh, wait, first paragraph says they're looking for both jr and sr Feb 26 16:52:51 24k? in London? they must be looking for a school leaver, right? Feb 26 16:53:12 ScaredyCat: prices aren't what they used to be Feb 26 16:53:30 I *started* work in London for more than that! Feb 26 16:53:50 that was a while ago now... Feb 26 16:53:53 what with the new EU countries and India-based labour, prices are getting squeezed Feb 26 16:54:02 and accounting for inflation... ouch! Feb 26 16:54:12 yeah, 24k is nothing Feb 26 16:54:46 <1300 take home? Feb 26 16:54:58 then there's your train ticket.. Feb 26 16:55:17 (cos there's no way you'll live near london at that salary) Feb 26 16:55:59 you be better off livin in india and working from there ;) Feb 26 17:07:56 ScaredyCat: I think that's probably the intention of the posting. They have no intention of hiring locally, but it's bad PR to make it obvious. Feb 26 17:10:36 ScaredyCat: aye, I think to live anywhere inwards of Zone 3 you need around 30k or a combined income Feb 26 17:11:07 hi...is it possible to build a rootfs with the mokomakefile without building the kernel? Feb 26 17:12:42 freerunner is looking good so far Feb 26 17:14:02 zefanja: No, the rootfs includes the kernel modules, so the current recipes have a dependency on the kernel. Feb 26 17:14:23 hrrump Feb 26 17:14:27 * ScaredyCat slaps XorA Feb 26 17:14:28 mwester-road: hmm...ok Feb 26 17:14:33 stop teasing us Feb 26 17:14:45 tease me some more! Feb 26 17:14:47 ScaredyCat: you guys wont have much longer to wait hopefully Feb 26 17:15:16 * mwester-road wants his GTA02 to work :p :D Feb 26 17:15:22 s/2/1/ Feb 26 17:15:23 if it's longer than 10 minutes it's too long Feb 26 17:15:24 sigh Feb 26 17:15:42 * mwester-road goes back to work, since he obviously can't type! ;) Feb 26 17:16:03 are you *on* the road... or en route Feb 26 17:16:19 ScaredyCat: well its better than getting a broken neo which would make you cry :-) Feb 26 17:16:39 true Feb 26 17:17:30 ScaredyCat: I'm travelling, but I'm not actually in motion. Perhaps I need to change the nick to better express my state... Feb 26 17:18:33 * XorA assumed it was road the company Feb 26 17:18:35 mwester-remote? mwester-not-at-home? mwester-away-from-his-neo-and-toys? :p Feb 26 17:18:49 mwester-I'm-travelling-but-I'm-not-actually-in-motion Feb 26 17:19:14 mwester-visting-a-customer-and-supposed-to-be-doing-real-work? Feb 26 17:19:59 Actually, customer is at "lunch" and playing cards in the cafeteria, so it's not so bad... Feb 26 17:25:13 XorA: any video ? Feb 26 17:25:22 of sexy stuff on FR Feb 26 17:26:24 ScaredyCat: I dont have a camera Feb 26 17:26:41 send me your freerunner :) Feb 26 17:26:45 I'll do it Feb 26 17:26:47 :P Feb 26 17:26:48 ScaredyCat: you should have come to FOSDEM Feb 26 17:27:32 If I could I would, but I'm already pushing my luck with 2 weeks in .nl and 2 in uk ...:( Feb 26 17:27:32 eeem Feb 26 17:27:36 I would if I could Feb 26 17:29:12 you have a working unit though? Feb 26 17:30:20 ScaredyCat: mostly working Feb 26 17:33:35 good, good... wifi ok ? Feb 26 17:33:39 stable? Feb 26 17:33:47 even under cpu load? Feb 26 17:34:33 I havent kicked its arse yet :-) Feb 26 17:36:47 :) Feb 26 17:37:31 has the price changed at all with the mods that have happened ? Feb 26 17:37:38 (end user price) Feb 26 17:39:24 no idea Feb 26 17:44:29 ScaredyCat: no good talking to me about prices and stuff, I ignore that part of the business Feb 26 17:45:13 Aren't you the liason ? Feb 26 17:45:16 450$ Feb 26 17:45:23 ScaredyCat: nope, Im OE developet Feb 26 17:45:30 s/pet/per/ Feb 26 17:45:31 XorA meant: ScaredyCat: nope, Im OE developer Feb 26 17:45:56 ScaredyCat: openmoko-distro is my game :-) Feb 26 17:46:01 mmmm... I'm sure you were something like that... Feb 26 17:46:10 ScaredyCat: Michael Shiloh is liason Feb 26 17:46:29 dagnabit! Feb 26 17:47:02 * ScaredyCat gets even more confused without coffee Feb 26 17:47:13 bitbake coffee Feb 26 17:47:29 -c rebuild Feb 26 17:47:36 ScaredyCat: http://www.openembedded.org/openmoko-hires-developer-devoted-to-openembedded Feb 26 17:47:46 Is there a special API I should use to talk to addressbook.db on OM, or is it the normal libebook? Feb 26 17:48:22 I prefer to use python or some other scripting interface if possible Feb 26 17:48:33 !!! Feb 26 17:48:39 wurp2: its normal eds files Feb 26 17:48:46 * ScaredyCat gets the whip out... Feb 26 17:48:59 I'm writing something to import sim card contacts into the addressbook Feb 26 17:49:11 so, was there ever the official announcement? Feb 26 17:49:39 * ScaredyCat wonders why all the python... Feb 26 17:49:42 ScaredyCat: app software shouldn't be written in C Feb 26 17:49:53 wtf! Feb 26 17:49:54 python pwns perios Feb 26 17:49:55 Or, rather, you shouldn't use app software written in C until it's had years to mature Feb 26 17:49:59 *period Feb 26 17:50:20 perios --- p shaped cherios Feb 26 17:50:30 as opposed to oerios Feb 26 17:50:33 I beg to differ... Feb 26 17:50:35 o shaped cheerios Feb 26 17:50:37 for embedded devices it should be written in C Feb 26 17:50:42 on the don;t use c ... Feb 26 17:50:47 pezackly Feb 26 17:50:57 I disagree Feb 26 17:50:57 python is extremely slow compared to C Feb 26 17:51:01 lua may be a better choice Feb 26 17:51:04 python is for ladyboys anyway Feb 26 17:51:06 wurp2, I got the fast charger installed.. The triple confirmation annoys me, but I can charge in my car now. :D Feb 26 17:51:19 lua isn't muh better, and it's not actually stable Feb 26 17:51:26 * mjr wiggles his bum and frowns at ScaredyCat Feb 26 17:51:30 Vegar: I defy you to find ppl who will care about the speed difference between C and python for importing their contact list, on a 200mhz machine Feb 26 17:51:33 Vegar: I disagree. IMO C++ should be used for embedded development too. Feb 26 17:52:03 * Stephmw whips out his ePeen to play with the BigBoys Feb 26 17:52:03 wurp2: it's not jst speed... it's deps Feb 26 17:52:03 kdean06: You mean the quadruple confirmation? Feb 26 17:52:03 :-) Feb 26 17:52:06 ScaredyCat: yes, back in october on community list Feb 26 17:52:08 Python is 16 times slower than C++, so it is usually pretty much out of the question (except maybe for very high level code). Feb 26 17:52:13 wurp2, Yeah. Feb 26 17:52:17 It may use less CPU processing :P Feb 26 17:52:19 * ScaredyCat stabs mjr in the bum with a ptch fork Feb 26 17:52:26 <_diego_> speed in 90% is not a problem Feb 26 17:52:27 kdean06: I can make that optional Feb 26 17:52:30 Tronic: I agree on that Feb 26 17:52:33 XorA: oh, on list... Feb 26 17:52:43 I understand not liking it, but I *really* wanted to make sure I'm not responsible for people frying hardware Feb 26 17:52:52 High speed isn't important. It's just got to be 'fast enough'. Feb 26 17:52:58 Vegar: Good, because now you can make a point of C++ development being just as easy as Python development :) Feb 26 17:52:59 wurp2, Don't worry about it. Feb 26 17:53:03 ScaredyCat: http://lists.openmoko.org/pipermail/community/2007-October/010731.html Feb 26 17:53:06 anyway, it should be rather obvious that the best choice of language depends on the particular problem; C (or, heaven forbid, even ++) does have a place with limited devices, but so does python Feb 26 17:53:07 Agreeing with Stephmw... Feb 26 17:53:17 and _diego_ Feb 26 17:53:27 ScaredyCat: but I was also the guy who wrote the GTA01 audio driver before the project was public Feb 26 17:53:32 as for the python bashing... there's already handsets shipping in Taiwan with python as the underlying app language Feb 26 17:53:38 <_diego_> i writing simple apps for neo in python and they are already fast enough ^^ Feb 26 17:53:40 Vegar: Otherwise you would have nothing to say when someone justified the use of Python by reduced development time and increased safety. Feb 26 17:53:48 It doesn't matter if it's 16 x slower (although I disagree with that assessment, once pypy is available) Feb 26 17:53:55 XorA: ahh ok, so you can actually code then :P Feb 26 17:53:58 if the end user can't tell any real difference in speed Feb 26 17:54:03 ScaredyCat: allegedly Feb 26 17:54:07 :) Feb 26 17:54:08 wurp2: right Feb 26 17:54:24 ScaredyCat: Im currently kicking OE into freerunner shape Feb 26 17:54:24 pypy in the skysky Feb 26 17:54:28 mjr: I agree completely Feb 26 17:54:37 excellent... Feb 26 17:54:54 I just think that for end user apps, the right choice is almost always java/python/ruby/... Feb 26 17:55:02 oh god no Feb 26 17:55:03 unless a very mature C/C++ alternative is available Feb 26 17:55:26 what about vala? Feb 26 17:55:38 wait... so your 1st gen python code is better than anyone elses 1 st gen C ? Feb 26 17:55:39 wurp2: Strange language grouping you use. Feb 26 17:55:45 the point on a handset is that given a suitably good runtime library, no app should be able to crash the handset short of a DoS Feb 26 17:56:09 <_diego_> "1st gen"? Feb 26 17:56:17 Tronic: Why do you say that? Feb 26 17:56:35 wurp2: Java and Python are in the opposite ends of the spectrum (C++ being somewhere in the middle), when it comes to strictness. C and C++ have almost nothing in common, when it comes to development style. Feb 26 17:56:40 _diego_: code just written Feb 26 17:56:48 Those are languages that tell you what went wrong when they crash, and virtually never corrupt memory and just go crazy Feb 26 17:57:04 and are not vulnerable (in general) to buffer overruns & the like Feb 26 17:57:26 Tronic: But python & java both do the same thing when you have bad code or crazy input Feb 26 17:57:47 http://www.google.com/search?q=python+buffer+overflow&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-GB:official&client=firefox-a Feb 26 17:58:03 wurp2: Use a few basic coding guidelines (that many just refer as modern C++) and you cannot have memory leaks, buffer overruns or other such issues in C++ either. In C you always get a lot of those, and they are not easily avoidable. Feb 26 17:58:24 Tronic: sorry, gotta call you on that Feb 26 17:58:34 yeah, I use strcpy all the time ... just for the danger :P Feb 26 17:58:35 Tronic: who's going to enforce those guidelines on 3rd-party apps? Feb 26 17:58:46 <_diego_> ScaredyCat: btw making mistake in C that causes overflow is much more easy to do in C than in Java or Python Feb 26 17:58:58 wurp2: In Java all objects are used via pointers, so in that respect it is quite similar to C. Feb 26 17:59:20 Tronic: They are not C style pointers, not even close Feb 26 17:59:29 <_diego_> Tronic: not at all Feb 26 17:59:30 Tronic: no they're not Feb 26 17:59:37 heh Feb 26 17:59:39 _diego_: those errors are easy to avoid if you use common sense and know what you're doing Feb 26 17:59:39 I agree _diego_, but that's down to the developer and the methods they use.. not the language Feb 26 17:59:44 wurp2: The only difference to C is that you don't have pointer arithmetics in Java. Feb 26 17:59:49 wurp2: And the syntax is nicer... Feb 26 17:59:57 and null pointers don't just flake out Feb 26 18:00:03 and you can't delete something twice Feb 26 18:00:07 Tronic: if you think Java references are pointers you clearly haven't done enough Java Feb 26 18:00:10 C++ can implement safe pointers in same style as java quite easilly Feb 26 18:00:16 and you can't use an uninitialized pointer Feb 26 18:00:18 in that case we need visual basic on the neo.... Feb 26 18:00:23 The use of a NULL pointer in C very reliably crashes the program on any modern CPU. Feb 26 18:00:27 and the lack of pointer arithmetic is a Big Deal Feb 26 18:00:34 XorA: agreed - my point is that it's impossible to force 3rd-parties to follow those rules Feb 26 18:00:36 In Java you get NullPointerException. Feb 26 18:00:38 <_diego_> ScaredyCat: i've worked with a lot of developer, and happens to meet people that don't use pointer couse they are difficult to understand Feb 26 18:00:43 Stephmw: look under the bonnet, Java stores its Reference as a void * pointing at an array that starts with the length of the array :-) Feb 26 18:00:44 Not much difference. Feb 26 18:00:45 XorA: once you let them go native, the box is theirs Feb 26 18:00:52 XorA: it's VM dependent Feb 26 18:01:01 Tronic: That is a vast difference Feb 26 18:01:02 _diego_: huh... they're not that difficult.... Feb 26 18:01:04 XorA: they're handles to a structure, sure, but not pointers Feb 26 18:01:10 where can i get the openmoko logo and fonts for download Feb 26 18:01:11 In C, it is common to write to the wrong place in memory Feb 26 18:01:12 Stephmw: Even the creators of Java call them pointers. Feb 26 18:01:16 they might need explaining, but it's not difficult Feb 26 18:01:17 <_diego_> ScaredyCat: so it's a problem of developer AND of language Feb 26 18:01:18 In Java, it is virtually impossible Feb 26 18:01:23 Tronic: and it's still wrong Feb 26 18:01:31 Stephmw: And they essentially are pointers (memory addresses). Feb 26 18:02:01 wurp2: wtf? Feb 26 18:02:03 The behavior exactly matches that of C pointers, except that there are no stars and no arrays (pointer arithmetic). Feb 26 18:02:08 Tronic: not the ones you see in Java... hashCode is most definitely *not* a pointer Feb 26 18:02:20 <_diego_> Tronic: they are not memory addresses, have u heard about Garbage collection? Feb 26 18:02:47 _diego_: Have you heard that GC can be use with C or C++ too? But how is this relevant? Feb 26 18:02:57 * Stephmw gives up on Tronic Feb 26 18:03:07 Stephmw: How come? Feb 26 18:03:08 anyway is time for me food Feb 26 18:03:09 XorA: do you see what I mean about enforcement though? Feb 26 18:03:10 * ScaredyCat puts Stephmw back in his box Feb 26 18:03:46 * ScaredyCat has hunger... Feb 26 18:03:46 Tronic: having worked on embedded VMs I'm basing my arguments on working experience Feb 26 18:03:51 _diego_: Look up Boehm GC (there are others as well). Feb 26 18:03:54 Tronic: but you're welcome to your opinions Feb 26 18:04:02 In C, when you have a buffer overflow (which is easy to do; just forget to validate one argument, or do it improperly), you have a security risk Feb 26 18:04:02 <_diego_> yes, if they are pointer to a memory address they should be changed every time a gc happens, but this not happens couse they are just id Feb 26 18:04:15 In Java, you have an exception with a stack trace pointing you to the error Feb 26 18:04:25 Stephmw: yes, I got no problems with Neo programs in type safe languages, its certainly sensible these days Feb 26 18:04:27 Stephmw: Does the Java object variable not contain a memory address of an object (even if a virtual address)? Feb 26 18:04:33 I have actually done Java development professionally for about 7 years Feb 26 18:04:35 <_diego_> that's the reason you can't even think to do aritmetics on it Feb 26 18:04:37 wurp2: that makes you bad at coding C nothing more Feb 26 18:04:41 Tronic: nothing accessible from Java Feb 26 18:04:41 Stephmw: we have a 500Mhz PC class machine, not a 7Mhz Amiga Feb 26 18:04:41 and C/C++ dev professionally for the 6 years prior Feb 26 18:04:50 XorA|gone: :) Feb 26 18:05:03 ScaredyCat: Totally wrong. Feb 26 18:05:04 Tronic: the VM hides all this Feb 26 18:05:11 Stephmw: Yes, I just stated that pointer arithmetics are not available in Java. That doesn't change the basic principle. Feb 26 18:05:17 ScaredyCat: If that were true, we would never find buffer overflows in big name apps Feb 26 18:05:43 Tronic: in any case, the 4-byte handle you refer to isn't a direct pointer in most VMs, it's a handle to a hash Feb 26 18:05:49 wurp2: I can tell that it was C/C++ and not C and C++ :P Feb 26 18:06:05 I dunno what you mean Feb 26 18:06:16 Sometimes it was pure C Feb 26 18:06:20 sometimes pure C++ Feb 26 18:06:22 sometimes a mix Feb 26 18:06:32 wurp2: C/C++ being something that is not modern C++, but rather C-like code that just uses classes. Feb 26 18:06:40 * Stephmw steps away for 5 Feb 26 18:06:40 as XorA said, it depends on the project Feb 26 18:06:40 Why did the topic get changed to an April release? I don't see any changes in the Community Updates page. Feb 26 18:07:05 bipolar: Some people in here thought March was very unlikely Feb 26 18:07:09 Others of us are less sure Feb 26 18:07:25 wurp2: ic Feb 26 18:07:57 <_diego_> bipolar: i think, if exit in march everyone is happy, if in april nobody complains :) Feb 26 18:08:13 hah Feb 26 18:09:13 bipolar: the community updates page doesn't mention March either Feb 26 18:10:26 wurp2: huh? Feb 26 18:10:46 bad practice is bad practice... Feb 26 18:11:28 march is very soon Feb 26 18:12:25 is the build of openmoko-tasks2 only on my system broken or does anybody also have problems with it? Feb 26 18:19:15 have you updated recently RedDog? Feb 26 18:19:51 i run make update; make setup; make openmoko-deve-image Feb 26 18:19:53 +l Feb 26 18:20:06 and this broken build is there for some days allready ... Feb 26 18:20:22 just going to try here... Feb 26 18:20:37 mm... mine built the othe ay ok Feb 26 18:20:57 s/othe ay ok/other day ok/ Feb 26 18:20:57 ScaredyCat meant: mm... mine built the other day ok Feb 26 18:22:58 ScaredyCat: are you using openembedded or openmokos mtn tree? Feb 26 18:23:33 oe Feb 26 18:23:57 okay, than i will give it a try :D Feb 26 18:30:09 ScaredyCat: I agree that bad practice is bad practice. It is also true that if you can systematically eliminate the possibility of a bad practice altogether, that's a big gain. Feb 26 18:30:35 ScaredyCat: It doesn't fix any buffer overflows to say "oh, the people who do that are just all bad coders" Feb 26 18:31:20 Switching to a platform in which buffer overflows are impossible demonstrably DOES fix buffer overflows Feb 26 18:33:02 yes, but pyhton isn't imume to buffer overflows... Feb 26 18:33:07 imune Feb 26 18:33:10 immune Feb 26 18:33:13 mm... Feb 26 18:33:18 mmm... Feb 26 18:33:39 and the first one that thinks python apps are good for a phone, will be taken out back and shot by me personally. Feb 26 18:33:45 shall we throw in the term 'resilient' in there? Feb 26 18:33:45 we dont need java 2.0 Feb 26 18:33:52 ~spel chicken Feb 26 18:33:55 ~spell chicken Feb 26 18:33:57 'chicken' may be spelled correctly Feb 26 18:34:02 xcasex: why bother suggesting it? it's an accomplished fact Feb 26 18:34:05 ScaredyCat: it isn't? Feb 26 18:34:05 ~spell immune Feb 26 18:34:09 'immune' may be spelled correctly Feb 26 18:34:16 xcasex: python apps are fine for a phone, depending on the app Feb 26 18:35:06 so python gained resilience and reliability. interesting. we had some awesome issues with the reliability part when writing eve. Feb 26 18:35:07 wurp2: http://www.google.com/search?q=python+buffer+overflow Feb 26 18:35:24 still do. Feb 26 18:36:37 we're talking about the game here, right? Feb 26 18:36:41 ScaredyCat: Right. Python, the VM, has buffer overflows in it Feb 26 18:36:46 Because it's written in C Feb 26 18:36:53 Stephmw, yeah. Feb 26 18:37:06 But it is one app, that when fixed, fixes that buffer overflow in all python apps Feb 26 18:37:08 wurp2, inheritance is a bitch, i agree Feb 26 18:37:38 xcasex: I didn't realise any part of it used python - is it all in it, or a subsection? Feb 26 18:37:41 xcasex: whoosh Feb 26 18:37:51 I didn't understand that remark :-) Feb 26 18:37:59 Stephmw, stackless python, the whole dang client. Feb 26 18:38:10 stackless python? Feb 26 18:38:18 wurp2: so python still suffers... Feb 26 18:38:22 Does that mean what it sounds like it means? I don't see how it could... Feb 26 18:38:29 yes. Feb 26 18:38:49 http://en.wikipedia.org/wiki/Stackless_Python Feb 26 18:39:01 37 bb's left RedDog Feb 26 18:39:01 ScaredyCat: That's like saying because .01% of the time the safety fails and people shoot themselves with the safety on, that safeties are bad Feb 26 18:39:37 ScaredyCat: 328/1.180 certs Feb 26 18:39:39 no... it's like saying 'Python never suffers from buffer overflows" Feb 26 18:39:43 OK, it's not stackless; it just implements its own stack Feb 26 18:40:06 xcasex: that's an interesting choice - I wish I had more info to give you on phones based on python Feb 26 18:40:08 there's quite a few updates RedDog Feb 26 18:40:24 xcasex: I would imagine that to start with they're less complex than a game ;) Feb 26 18:40:42 (the apps individually that is) Feb 26 18:40:49 and yet, the need for reliability is far greater Stephmw Feb 26 18:41:22 and i wouldnt buy a phone that touted that core functionality was implemented in python. Feb 26 18:41:29 s/in/with Feb 26 18:41:37 xcasex: this is one I was thinking of, http://www.linuxdevices.com/news/NS5617889988.html Feb 26 18:41:43 I would... if I could ter it out :_ Feb 26 18:41:46 tear Feb 26 18:41:59 xcasex: I'm a J2ME guy, can you break down the lack of reliability? Feb 26 18:42:03 (in python) Feb 26 18:42:27 It sounds to me as if the python implementation they chose is not broadly enough used to have been made reliable yet Feb 26 18:43:27 wurp2, yes because 23/7 (hour down due to backup of sql cluster) uptime with 27k concurrent users isnt a broad enough crowd. Feb 26 18:43:31 ScaredyCat: My point is that there is a huge, categorical difference between the frequency of buffer overflows impacting python & java versus a C app Feb 26 18:44:27 xcasex: The userbase that matters for the reliability of a VM/compiler implementation is the # of devs working on it & reporting bugs, not the # of people running programs on it Feb 26 18:44:59 Stephmw, you dont want to peg the cpu with an interpreted language, jit is fine since its bytecode in one way or the other (although i wouldnt recommend it); this is all imho btw, but. i want reliability which means resources for the sw to do its job. Feb 26 18:45:09 wurp2, what do you think #users gives #devs Feb 26 18:45:15 bugreports. Feb 26 18:45:28 ScaredyCat: and furthermore, over time the python & java VM/compiler will get get more reliable, whereas each C program mostly starts from the same unreliable starting place Feb 26 18:45:48 what is this? invasion of the non-comprehension? Feb 26 18:46:05 xcasex: So your users give you bug reports that indicate how the VM failed, then you guys go fix it? Feb 26 18:46:11 xcasex: ok, I'll have to accept those for python. Feb 26 18:46:15 I don't understand how it can still be unreliable, then. Feb 26 18:46:50 wurp2, correct. we ship it with debugging enabled and a debugging tool. so yes. heh. heavy information density in those bug reports. Feb 26 18:47:48 xcasex: the VMs I've worked on have always been on RTOS, so they couldn't DoS the phone's CPU Feb 26 18:48:08 xcasex: the watchdog would just cycle the app if it was a system one, or just kill it if it was 3rd-party Feb 26 18:49:12 xcasex: Are your reliability problems in your python code, or in the VM? Feb 26 18:49:56 Stephmw, python is dog slow on a p3 400; why would it fare better on a 200mhz risc (ironically risc is a misnomer nowadays) Feb 26 18:50:05 wurp2, hahaha. are you arguing for the sake of it? Feb 26 18:50:25 ?? Feb 26 18:50:36 <_diego_> xcasex: depends in what are you doing Feb 26 18:50:36 I was asking you a question... Feb 26 18:50:48 So I could understand the problem. Feb 26 18:51:19 <_diego_> for handling a gui on 200Mhz cpu is enough (libraries are written i C) Feb 26 18:51:22 xcasex: nowdays on current 200MHz arm cores, the display hw is usually the bottleneck with java apps... but considering non-speed critical apps like media/video/audio (that can be done with native apis), UI interaction only needs to be fast enough anyway Feb 26 18:51:36 Generally I find it better to take a position based on information, rather than unfounded assertion Feb 26 18:51:52 wurp2, the problem is always in the code, wherever it is. vm, server, client etc. Feb 26 18:52:02 xcasex: ... ok, I forgot a 'not' in there ;) Feb 26 18:52:03 fix one bug, open up another. Feb 26 18:52:07 xcasex: you get the idea Feb 26 18:52:12 Stephmw, im aware of this :) Feb 26 18:52:49 xcasex: But if your argument is that "python is unreliable" Feb 26 18:53:14 it matters whether your reliability issues are in the python code or in the VM Feb 26 18:53:27 and whether the VM is a mainstream one or a sideline Feb 26 18:53:38 wurp2, my argument is i'd rather trust java. full blown java. before i see python. Feb 26 18:53:49 hell. tomcat the phone. Feb 26 18:53:56 <_diego_> why? Feb 26 18:54:07 Java is very reliable Feb 26 18:54:18 I would absolutely trust java Feb 26 18:54:29 I wouldn't run it on a phone right now, because the memory requirements are way too high Feb 26 18:54:36 anyone know if GTA02 is going to have EDGE? I presume zero chance of it doing 3G. Feb 26 18:54:48 pjz: No EDGE Feb 26 18:54:49 and those mem requirements equally apply to interpreted languages. Feb 26 18:54:51 pjz, no. Feb 26 18:54:54 pjz: Wifi + gprs Feb 26 18:55:04 <_diego_> xcasex: no Feb 26 18:55:05 wurp2: which java? ME or SE? Feb 26 18:55:13 xcasex: Not true. Different interpreted languages have very different memory characteristics Feb 26 18:55:26 so. why python and no boo or lua? Feb 26 18:55:27 Stephmw: Sorry; I don't know much about J2ME Feb 26 18:55:27 <_diego_> different language have different memory requirement Feb 26 18:55:36 _diego_, also true. Feb 26 18:55:37 hmm, I ran java on an 8 bit uC with only 4k of RAM, java isnt that bad :-) Feb 26 18:55:48 XorA|gone, hehehehe Feb 26 18:55:52 I would tend to trust j2me's reliability a lot less, since I thought it gets a custom implementation on each platform Feb 26 18:55:54 XorA|gone: for some value of 'java' Feb 26 18:55:56 wurp2: average config is for a 2MB heap max, 512KB min Feb 26 18:56:01 s/custom/basically custom/ Feb 26 18:56:01 wurp2 meant: I would tend to trust j2me's reliability a lot less, since I thought it gets a basically custom implementation on each platform Feb 26 18:56:37 s/hehehehe/testing this fly shit out/ Feb 26 18:56:39 xcasex meant: XorA|gone, testing this fly shit out Feb 26 18:56:39 wurp2: that last point is definitely true... the JSRs for J2ME have too many optional aspects and the wording is woolly Feb 26 18:56:44 oh nice. Feb 26 18:56:52 Stephmw, i bet. Feb 26 18:56:53 wurp2: *but* Feb 26 18:57:19 wurp2: the testing done is very thorough, the TCKs aren't just glanced at :) Feb 26 18:57:26 wurp2: but tests can be contested Feb 26 18:57:51 <_diego_> probably tcl/tk, have lower memory requirement... :) Feb 26 18:57:58 Stephmw, is the j2me certs as rigid as the j2ee? Feb 26 18:58:07 <_diego_> even of C/gtk apps Feb 26 18:58:09 javacard TCK was a bitch to pass Feb 26 18:58:09 xcasex: for developers? more so Feb 26 18:58:20 Stephmw, interesting. Feb 26 18:58:33 xcasex: for VM implementors... it's probably similar and more rigorous Feb 26 18:58:58 xcasex: the dev cert requires knowing a fuckload of JSRs intimately Feb 26 18:59:00 Stephmw, which was what i presumed. Feb 26 19:00:36 xcasex: it's a shame really... I could write a blog fulltime after having worked... where I have worked... but I'd be in breach of the various NDAs :( Feb 26 19:00:53 xcasex: even the TCKs have em Feb 26 19:02:36 wurp2, i still -- despite the pm you sent -- consider python to be a bad idea for the freerunner. especially because it is python. although using python to test embedded firmware etc. is all well and fine. or if you're doing something similar to the XO Feb 26 19:02:59 but hell if we're so keen on python, Rewrite the kernel in it. Feb 26 19:03:14 <_diego_> python is bad becouse is python?? Do u call it an explanation Feb 26 19:03:36 xcasex: oh god no, not a typesafe kernel... I've seen that done in Java Feb 26 19:03:45 Stephmw, oh. i know. i was about to headbutt Erikssons j2ee group when they presented glassfish at osscon in stockholm. Feb 26 19:03:46 xcasex: it's a nice research idea, but totally unworkable ;) Feb 26 19:04:05 <_diego_> so, your argoument, i'm right becuase i'm xcasex? :D Feb 26 19:04:12 _diego_, python is bad, for a resource sparse system, because it is a interpreted language. Feb 26 19:04:22 _diego_, yes. normally i'd settle for that ;) Feb 26 19:04:37 <_diego_> 128M are a LOOOOT of ram :) Feb 26 19:04:56 xcasex: I don't understand why you are flipping to this extreme straw man argument Feb 26 19:05:02 but doing voip service bridges for telecom services is far more interesting ;) Feb 26 19:05:03 <_diego_> really, i've written apps running on 32Mb Feb 26 19:05:05 and I agree with _diego_ Feb 26 19:05:06 wurp2, tssk. Feb 26 19:05:33 128mb is a lot of ram. i agree. but then you have to factor in system requirements, overhead etc. Feb 26 19:05:36 I am proposing that python is good for end user apps Feb 26 19:05:43 which it is Feb 26 19:05:47 oh i agree. Feb 26 19:05:52 but not on a phone. Feb 26 19:05:58 not that it is good for kernel, or device drivers, or any other straw man you happen to come up with Feb 26 19:06:00 a lot of people will be able to script their programs in python.. Feb 26 19:06:05 5 years ago, absolutely not on a phone Feb 26 19:06:13 today, yes on a phone Feb 26 19:06:29 <_diego_> and what about Commodor Vic20, have an interpreted language and 4k or ram Feb 26 19:06:48 <_diego_> s/or/of/ Feb 26 19:06:48 _diego_ meant: and what about Commodof Vic20, have an interpreted language and 4k or ram Feb 26 19:06:52 I wrote assembler code on C64, because it ran at 4khz and had 64 k of ram Feb 26 19:07:00 I think python is suitable for the neo, it has enough resources Feb 26 19:07:25 the neo runs 50,000 times faster and has 16000 times more ram Feb 26 19:07:28 edistar_, a bit offtopic, yellowtab edistar? Feb 26 19:07:47 wurp2, i'm arguing reliability and resilience. Feb 26 19:08:10 A minute ago you were arguing "not for a phone" Feb 26 19:08:16 <_diego_> xcasex: _diego_, python is bad, for a resource sparse system, because Feb 26 19:08:17 <_diego_> it is a interpreted language. Feb 26 19:08:18 guarantee a uptime that a normal closed source proprietary app has. -- jokes aside -- and i'll buy the argument Feb 26 19:08:23 Is there some way reliability and resilience in python suffer when it's running on a phone? Feb 26 19:08:33 _diego_, that is part of the argument. Feb 26 19:09:14 wurp2, yes. Feb 26 19:09:17 in short. Feb 26 19:09:23 <_diego_> hahahaha Feb 26 19:09:41 xcasex: Very short - an assertion with no data or logic listed to back it up. Feb 26 19:10:02 wurp2, well then. show me a python app that doesnt crash, and i'll buy the argument ;D Feb 26 19:10:22 my website at www.moochmuch.com Feb 26 19:10:35 my fast charge checking daemon on the neo Feb 26 19:10:45 zope servers Feb 26 19:11:08 I assume by "doesn't crash" you mean "crashes more frequently than a similar C app" Feb 26 19:11:08 website software on a server. with gobs of ram? Feb 26 19:11:11 ... Feb 26 19:11:14 yes. Feb 26 19:11:25 although honestly I can't think of a time any of those have crashed period Feb 26 19:11:51 The fast charge checker is running on the neo Feb 26 19:12:08 although, quite frankly, that's one that I think would benefit from being ported to C Feb 26 19:12:12 since it stays resident Feb 26 19:12:29 but, it doesn't crash, or hasn't in the week or so I've been running it Feb 26 19:12:29 bah. Feb 26 19:12:48 xcasex: very eloquent Feb 26 19:13:24 rewrite the call manager in python, its more complex than an app polling /proc. Feb 26 19:13:56 neat Feb 26 19:14:08 You keep making assertions with nothing to back them up, then responding with sidesteps when I back up mine. Feb 26 19:14:29 is the gta01 suspend2ram now stable with fix-i2c-s3c2410-resume-race.patch? Feb 26 19:14:55 RedDog: Everything I've heard indicates it is, although some people don't have their phone wake when they receive calls Feb 26 19:15:21 I am not using it yet Feb 26 19:15:22 hmm ... sounds good! ;) Feb 26 19:16:34 wurp2, back it up, with more words you mean. look. let the code talk instead. Feb 26 19:16:42 I did Feb 26 19:16:53 You asked for any reliable python apps Feb 26 19:17:05 I listed three examples, two of which I wrote, one of them running on the neo Feb 26 19:17:35 It is a relatively simple app, but that shouldn't make much difference, since it runs 24/7, if python is inherently unreliable on the neo Feb 26 19:20:36 wurp2, the stack heap of the vm isnt constant is it? so more complexity larger segments of code, larger heap, more ram in use, and resource scarcity isnt something you look for in a embedded device. right? Feb 26 19:20:36 I'm going to stop spamming people with arguments the either will not change their mind about, or don't care, or both, so you won't hear any more about java vs python vs C vs whatever from me here today :-) Feb 26 19:20:46 oh hehehe thanks :) Feb 26 19:21:38 s/ the ei/they ei/ Feb 26 19:21:39 wurp2 meant: I'm going to stop spamming people with argumentsthey either will not change their mind about, or don't care, or both, so you won't hear any more about java vs python vs C vs whatever from me here today :-) Feb 26 19:30:41 RedDog: NOTE: build 200802261828: completed Feb 26 19:33:10 ScaredyCat: Dammit, I thought you were saying that there's a new daily build available Feb 26 19:33:32 wurp2, There will be one tonight. Feb 26 19:33:40 Or... Tomorrow, depednign on your time zone. Feb 26 19:33:45 Yay!! Feb 26 19:33:55 The guy who does the builds has been at FOSDEM. Feb 26 19:33:58 GMT-6 for me Feb 26 19:36:37 wurp2: my build will be up in 10 mins or so... Feb 26 19:51:41 wtf is avahi doing on here! Feb 26 19:52:25 whats wrong with avahi Feb 26 19:52:43 how about its complete lack of security Feb 26 19:56:44 fewd! Feb 26 19:59:52 ScaredyCat: mine had an error... did make force-rebuild, now i takes times :-( Feb 26 20:00:33 For future reference, it looks as if the right way to manipulate the addressbook on OM is dbus (answering my own question from earlier) Feb 26 20:02:14 you'd better use a shell script to avoid buffer overflows :P Feb 26 20:02:36 gonna :-) Feb 26 20:02:40 not shell script Feb 26 20:02:42 python Feb 26 20:02:52 RedDog: hrrm.. I'm runing mine on my phone Feb 26 20:04:21 ScaredyCat: seems that you have a faster pc ... Feb 26 20:07:50 I just rebuild a lot Feb 26 20:08:02 rather than once a day Feb 26 20:09:07 seems that werner's patch to remove aux from stating up from a suspend breaks the recovery completely Feb 26 20:09:25 my build i broken for about one week, now he rebuilds all ... Feb 26 20:10:14 ScaredyCat: wich reconvery, hope you don't mean the resume from suspend... Feb 26 20:10:46 i do Feb 26 20:12:59 $§")/&§$% ... Feb 26 20:13:07 Sounds like another build not to install... Feb 26 20:17:31 ScaredyCat: but why? only "enable_irq_wake(irq_aux);" is removed... Feb 26 20:31:06 ScaredyCat: Is it possible you are having some other spurious issue with resume from suspend? Feb 26 20:32:25 well, I can't get it to wake up any other way Feb 26 20:33:35 ScaredyCat: oic Feb 26 22:03:24 i've been playing around with getting sidplay to run on openmoko and it's working good - however, what audio device is best to use (it defaults to /dev/dsp - but when starting sidplay from terminal I get a device busy!??) Any ideas? Feb 26 22:04:34 subtwo: Does sidplay support PulseAudio output? Feb 26 22:06:02 If not then it's either using pasuspend (if available already), killing PulseAudio and using ALSA (not good) or using the PulseAudio ALSA wrapper (eats performance) Feb 26 22:06:04 well, I am no audio guru by any means and have just begun digging, but I guess that the /dev/dsp device is enough for it. It works if you start playing through an ssh session... Feb 26 22:06:48 It works then because PulseAudio frees /dev/dsp after about 20 seconds of not playing - it does however play the touchscreen click sound, so if you use the touchscreen to run it, /dev/dsp will be occupied Feb 26 22:07:54 Yes, I already though of this, and removed the touchscreen sound from /etc/pulse/session - but still same result... Does it allocate the audio device even though no sound is played? Feb 26 22:09:27 It shouldn't. Did you restart PulseAudio after the config change? Feb 26 22:09:37 I even rebooted... Feb 26 22:10:14 Maybe I was too quick to try it after the startup sound though... Feb 26 22:12:07 Seeing how the last release of sidplay was in 2004 it doesn't seem like we'll get PA support... Feb 26 22:13:56 No, what is actually needed to add support for PA? Feb 26 22:16:02 subtwo: Well, the audio playback mechanism would have to be changed to use PA instead of OSS Feb 26 22:16:34 Also, there's a sid plugin for gstreamer which you might want to try out, too - the Media Player should be able to play back sid files with it installed Feb 26 22:16:52 Strangely, it does not still run from the terminal on the hardware but very well from an ssh session Feb 26 22:17:11 does pulseaudio still hog cpu like crazy in recent builds, btw? Feb 26 22:17:20 I'll check out the sid plugin for gstreamer - it is probably the best bet Feb 26 22:17:21 mjr: Nope Feb 26 22:17:46 When I start sidplay through ssh it eats about 10% of CPU time to play Feb 26 22:17:49 mjr: A nice little find by someone I don't remember dropped usage quite a bit with a simple usleep()... Feb 26 22:18:24 abraxa_, ack. Sounds ... dirty, but if it works for now... Feb 26 22:19:14 abraxa_: how can I use the PA ALSA wrapper? Feb 26 22:20:53 subtwo: Well, I said that assuming that sidplay supports both ALSA and OSS. I'm not sure if it does so if it only supports OSS then that won't work, unfortunately. If you feel like trying: http://www.pulseaudio.org/wiki/PerfectSetup#ALSAApplications Feb 26 22:21:43 abraxa_: thanks, you're probably right. I will check it out. Feb 26 22:22:22 subtwo: There's also an OSS wrapper, though I'm not sure it's included in the OM images: http://www.pulseaudio.org/wiki/PerfectSetup#OSSApplications Feb 26 22:26:15 abraxa_: thanks, I'll check it out. I'm after a simple "dirty" solution for now. I'm going to do a small presentation of the OpenMoko and would like to show how to build some simple library and application (libsidplay and sidplayer) and run it from the terminal. Feb 26 22:26:42 Ahh nice =) Feb 26 22:26:48 Let us know how that goes, please Feb 26 22:28:52 Yes, sure will. I will probably just kill the PA and see if that helps. Feb 26 22:28:56 freesmartphone.org: 03emdete * r80 10/trunk/software/pyneog/ (10 files in 3 dirs): first drop of the pyneo-gui seen at the fosdem Feb 26 22:31:33 abraxa_: It seems like the interest in OpenMoko is quite high, but nobody really have any real life experience with it. I think from an educational point of view it is the perfect lab-board for trying out embedded programming. Feb 26 22:32:07 abraxa_: The FIC hardware that is... Feb 26 22:33:10 g'night all ! Feb 26 22:36:34 subtwo: It might be, though it already is abstracted a lot and behaves just like any other linux system anyway :) Feb 26 22:37:45 abraxa_: Well, it is not necessary to run linux ;-) Feb 26 22:37:55 True Feb 26 22:42:25 abraxa_: Thanks for the tips about PA. I think I will settle for killing the PA daemon altogether it is working like expected. Feb 26 22:42:56 Alrighty. Good luck with your endaevors! Feb 26 22:43:51 Thanks! Feb 26 22:48:38 freesmartphone.org: 03emdete * r81 10/trunk/software/pyneog/INSTALL: hint to deinstall moko-gtk apps added Feb 26 23:52:53 topic Feb 27 02:47:09 Anyone here got it running in canada? Feb 27 03:50:31 moo Feb 27 03:54:32 Hey guys. Feb 27 03:54:48 Has anything new and nifty been added in the last month or so? Feb 27 04:18:15 openmoko: 03werner * r4119 10/branches/src/target/kernel/2.6.24.x/patches/gta01-inputdevice.patch: Feb 27 04:18:15 openmoko: Oops, didn't mean to break GTA01Bv3. Feb 27 04:18:15 openmoko: gta01-inputdevice.patch: Feb 27 04:18:15 openmoko: - drivers/input/keyboard/neo1973kbd.c (neo1973kbd_probe): reverted change in Feb 27 04:18:15 openmoko: r4114 and made AUX a wakeup source for GTA01 again Feb 27 04:18:18 openmoko: - drivers/input/keyboard/neo1973kbd.c (neo1973kbd_probe): added a comment Feb 27 04:18:20 openmoko: explaining why Feb 27 04:36:26 Wow, I just tried uImage-2.6.24+svnr4114-r4110-r3-neo1973.bin and it doesn't boot... just prints the same thing over and over. Feb 27 04:40:37 RayDar: If you ssh into the Neo and start /etc/init.d/xnodm then it'll run X and should run it from then on as supposed to Feb 27 04:44:40 abraxa__: Thanks for the tip. I will give it a try. Feb 27 04:46:36 hallo mokomonkeys Feb 27 05:07:16 Hi Holy Feb 27 05:07:29 abraxa__: I don't have an /etc/init.d/xnodm Feb 27 05:08:32 abraxa__: Ah, I see /etc/init.d/xserver-nodm Feb 27 05:09:25 abraxa__: Hmmm, that didn't seem to fix it. Feb 27 05:10:21 abraxa__: Doh! I ran it again and it started up.... kinda flaky Feb 27 05:45:28 Heheh, I have a newer version of the kernel on my phone than on my laptop and desktops Feb 27 07:15:42 laterz dudes Feb 27 07:36:09 openmoko: 03werner * r4120 10/developers/werner/memspeed/ (. Makefile memspeed.c): Quick and dirty physical memory speed tester. Feb 27 07:37:18 moin Feb 27 07:38:31 morning Feb 27 08:01:47 good morning Feb 27 08:15:55 mornning Feb 27 08:18:04 bad werner, naughty werner - in your bed. Feb 27 08:27:51 morning Feb 27 08:41:10 hm.. gllin was working fine yesterday, now i get 'stack size 8860 bytes (9k)' and it quits. (Yes i do have a cat /tmp/nmeaNP running) Feb 27 08:57:31 moin all Feb 27 09:18:13 <_diego_> Psi_, sometimes happen to me too, but usually if i restart it, it works fine Feb 27 09:22:25 _diego_, i restarted once and it didnt seem to help Feb 27 09:59:02 maybe due april now what ? Feb 27 09:59:30 what changed ... Feb 27 10:41:38 freesmartphone.org: 03mickeyl * r82 10/trunk/software/pyneod/setup.py: pyneod: repair setup.py syntax Feb 27 10:45:13 can someone provide a recent MokoMakefile that builds fine? Feb 27 10:54:31 Did you try make update? Feb 27 10:56:50 I'm using the 20080220 image and I'm experiencing a problem... gsm stops working after suspend and doesn't work when the phone wakes up... Feb 27 10:57:13 I queried bugzilla, but I couldn't find any bug report on that Feb 27 10:57:27 is it a known issue? Feb 27 10:58:16 It's more a case of at this point, it's still not stable Feb 27 10:58:49 the hardware or the software? Feb 27 10:58:52 or both? Feb 27 10:58:54 The software Feb 27 10:59:19 there is limited testing of new versions before committing. **** ENDING LOGGING AT Wed Feb 27 10:59:57 2008