**** BEGIN LOGGING AT Fri May 04 02:59:56 2007 May 04 03:24:28 (script) wiki RecentChanges: [[Wishlist:Text_Input]] May 04 04:12:12 good morning May 04 05:23:28 (script) wiki RecentChanges: [[Wishlist:Text_Input]] May 04 05:25:34 * CM rubs eyes May 04 07:21:43 mickey|bbl: are yu here May 04 07:21:44 ? May 04 07:22:47 openmoko: 03thomas * r1906 10/trunk/src/target/OM-2007/artwork/icons/openmoko-standard/ (22x22/Makefile.am 36x36/Makefile.am): icons: add missing makefiles May 04 07:30:37 counter May 04 07:30:38 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, 3 days 16:29:20 (10.687 +-3.5 days) (1085;193) May 04 07:32:12 counter May 04 07:32:13 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, 3 days 16:27:45 (10.686 +-3.5 days) (1086;193) May 04 07:32:56 openmoko: 03thomas * r1907 10/trunk/src/target/OM-2007/artwork/icons/openmoko-standard/ (22x22/stock/Makefile.am 36x36/stock/Makefile.am): icons: add missing makefiles May 04 08:22:09 omg the fpga resurfaces May 04 08:23:28 damn, you beat me to it May 04 08:24:05 morning May 04 08:24:09 Morning. May 04 08:28:22 buz, so it has been said before ;) May 04 08:29:02 not just once May 04 08:29:24 ok well. I'm following the list since a few weeks May 04 08:29:45 this particular use is quite sane to what was suggested before, though May 04 08:29:46 sorry to bring back the hell ^^ May 04 08:29:59 some people actually wanted to implement the whole phone in an fpga May 04 08:30:22 it remembers me a french company that did that for a pda May 04 08:30:59 if I remember correctly, they used 5 of them and were doing on-the-fly re-programmation depending the usage May 04 08:31:26 * Rince prefers umts instead of movieplaying May 04 08:31:31 ;) May 04 08:31:35 but it has never been so successful apparently :P May 04 08:31:45 battery will lsat not very long May 04 08:32:11 aesci7E: was it fast? May 04 08:32:26 I never tried the real device ^^ May 04 08:32:57 * aesci7E has no money to spend for a so so expensive pda (~800$ at that time) May 04 08:34:17 afaik, some of the 3ware raid controllers were in fact shipping on FPGA May 04 09:37:32 howdy May 04 09:44:34 Hi, everyone. May 04 09:45:00 I played a bit more with GPS decoding... it seems like incoming packet 3 contains bits gps is receiving. May 04 09:45:10 ...in "12 bits per received bit". May 04 09:46:08 Heh, 12 bits per bit.. May 04 09:47:29 CM: Like this: RECV L 7 SAT 8 FL 8 CMD 3 bc f ff d ff d 3f f fb e bb e 3d f 3a 0 fb 0 3f 1 7f 0 81 0 7f 1 May 04 09:47:30 3 1 c4 1 83 1 May 04 09:49:12 I hope someone can figure out how to translate it to coordinates May 04 09:49:23 morning May 04 09:50:35 CM: Translating it to bitstream from sattelites is the neccessary first step. May 04 09:50:58 Of course, and I'm happy to see the progress :) May 04 09:51:33 Hello, LaF0rge. May 04 09:51:37 hi pavel May 04 09:52:56 Trying to decode hammerhead GPS... but is there any chance we get useful code out of globalLocate (or how are they called?) May 04 09:53:38 pavel: we only have a proprietary library, which we cannot yet ship to customers/users May 04 09:53:50 pavel: i guess within two weeks we should be able to ship that library. no useful code. May 04 09:54:13 pavelm: do you have another gps unit? to compare result May 04 09:54:13 s May 04 09:54:37 hrw/pavel: the GL chip will not give you coordinates. it's a softcore gps May 04 09:54:39 hrw: Yes, I have bluetooth GPS here. May 04 09:54:50 hrw: the positioning information has to be calculated on the host. May 04 09:54:57 laf0rge: I know. I am trying to decode bitstreams it should be giving me. May 04 09:55:04 ok May 04 09:55:08 pavel: all the best May 04 09:55:20 laf0rge: (Sattelites broadcast something at 50bps, and those data should be useful for decoding position) May 04 09:55:29 (script) openmoko-devel: Werner Almesberger Finger text input idea May 04 09:55:30 pavel: to keep our relationship with GL clean, I unfortunately cannot work on a free software reimplementation, since I have sene the docs under nda May 04 09:55:44 seen May 04 09:56:07 laf0rge: So the "power up the gps" code in our kernels is "known good", because the binary library works with it? Good. May 04 09:57:06 openmoko: 03laforge * r1908 10/trunk/src/target/u-boot/patches/ (set-hwswp.patch uboot-s3c2410_fb.patch): * s3c2410fb: set the INVVCLK bit, just like the kernel does for quite some time May 04 09:57:13 pavel: it should work. there was one remaining bug, I would have to check with the GPS engineer here at FIC May 04 09:57:21 not sure if that one bug was already fixed. May 04 09:57:34 will check May 04 09:57:45 Aha, ok, I guess I'll just wait before trying anything on openmoko. May 04 09:57:57 I'd have to compile my own kernel first (etc), anyway, so.. May 04 10:01:09 openmoko: 03laforge * r1909 10/trunk/src/target/u-boot/patches/uboot-mokoversion.patch: bump version to moko8 to mark first Bv4 phase-0 release step (including i2c charger fix and INVVCLK fix May 04 10:06:43 hrw: which nokia battery is the one that files Neo? May 04 10:06:57 xora: The one from n6230 May 04 10:07:12 XorA: bl-5c May 04 10:07:20 6230, 1600, 2300, 3200 phones May 04 10:07:39 hrw: cool, my wife found a 1600 on the road yeaderday, a bit smashed, but it powers up May 04 10:08:24 xora: Try to return the sim card :_) May 04 10:08:50 XorA: remember that it will not charge neo battery May 04 10:09:15 but neo will work with 1600 battery May 04 10:09:27 pavelm: no sim or keypad May 04 10:09:45 pavelm: otherwise that SIM would be in Neo1973 making lots of GPRS data May 04 10:10:40 xora: :-) May 04 10:10:44 where is this world going to... May 04 10:13:34 XorA: Would you seriously take advantage of someone else's SIM? May 04 10:13:41 Can I use a normal lipo charger to charge the neo battery? (how many cells is the battery?) May 04 10:14:09 abraxa_: of course May 04 10:15:23 Of course?! Now that makes me wonder. May 04 10:18:40 abraxa_: wonder what? May 04 10:21:27 Why you would take advantage of someone who lost his/her cell by using the SIM on his/her expense May 04 10:42:55 sannes - You can just buy a bl5c charger. May 04 10:46:10 sannes: 1 May 04 10:46:53 sannes: Just charge at 300mA or so max, up to a maximum of 4.1V, and hold at 4.1 for an hour. May 04 10:47:03 +-0.05V May 04 10:50:31 Elrond: as long as it will be 3rdparty not nokia May 04 10:51:01 hrw - Yeah. May 04 10:51:16 If it's completely flat, then you can moderately safely connect a 3.3V power supply. May 04 10:51:24 Does charging in the neo now work better? May 04 10:52:05 Elrond: for me it is hard to tell for bv4 as I have bv3 only May 04 10:52:55 For bv3 it was a problem, if the battery is completely empty, right? May 04 10:53:01 yes. May 04 10:55:10 I hope that's better with v4. May 04 10:55:29 ... and it means, if one wants to chip bv3, one should load the battery full and then remove it. May 04 10:55:32 Some of it can be fixed in software. May 04 10:56:03 If you turn on the charger immediately in u-boot at 100mA, then immediately turn off and set a wakeup interrupt in 5 min. May 04 10:56:21 set a wakeup interrupt then turn off ratehr. May 04 10:59:08 Ahh, only affects standby time, but not current in "off"-status, right? May 04 11:01:57 IIRC, yes. May 04 11:05:14 Okay, so switching the phone of should be safe for chipping it. May 04 11:05:30 s/ of / off / May 04 11:05:30 Elrond meant: Okay, so switching the phone off should be safe for chipping it. May 04 11:12:54 chipping? May 04 11:24:29 (script) wiki RecentChanges: [[Wish_List]] May 04 11:34:33 shipping. ;) May 04 11:42:49 shipping? May 04 11:43:01 Oh. May 04 11:43:14 Personally. May 04 11:43:28 I'd recommend removing the battery. May 04 11:47:53 Yeah, removing is of couse safest. May 04 11:48:46 The nokia 3330 has a nice feature. May 04 11:48:57 You can put the battery in rotated 180 degrees. May 04 11:49:11 And it will lock in, and be not-switchable-on by accident. May 04 12:02:08 What we may consider a nice feature could very well become more load for the support hotlines ;) May 04 12:02:30 * abraxa_ wonders how many customers put the battery in wrongly and didn't get the cell to power up May 04 12:02:44 abraxa_: i do agree, my mum brought a phone and it wouldn't turn on, battery in the wrong way around May 04 12:02:52 Probably true. May 04 12:03:21 Personally, I like the option. May 04 12:03:25 SpeedEvil: i do still think it's a good idea, although would have to be well documented! May 04 12:03:34 Same here May 04 12:03:48 'absolutely positively off, and will still be charged in a month, nomatter what the software does, or accidental button presses'. May 04 12:04:11 quick question: whats the AUX button do? May 04 12:04:39 Stuff. May 04 12:04:44 Context dependant. May 04 12:04:56 * arw hello to all May 04 12:05:07 it is _not_ an emergency call button any more. May 04 12:05:14 At one point long ago, it was going to be. May 04 12:05:53 oh ok, nice.. my friend's just brought a k810i, another mate a w650i, me i'm holding out for the daddy Neo1973 May 04 12:06:18 abraxa: That "rotated battery" happened here, too :-) May 04 12:06:54 hehe May 04 12:07:16 webjames: Good decision :) May 04 12:07:44 I feel like such a fanboy May 04 12:07:46 abraxa_: thanks.. i know it will be May 04 12:08:04 If we get Neo keychains I'll definitely use that one instead of the one I have now May 04 12:08:14 And I wanna get an OpenMoko shirt as well... May 04 12:08:46 A working neo keychain with a touchscreen 0.5" display, talking via bluetooth would be _so_ cool. May 04 12:09:07 "working" neo keychain? May 04 12:09:34 Just a small neo shaped one, with a bluetooth display and touchscreen, maybe mic too. May 04 12:11:12 My bad - I mixed up keychain and lanyard May 04 12:11:27 I meant "if we get an OpenMoko lanyard with the Neo"... May 04 12:22:42 maybe the keyring could use the a-gps to locate you phone, emitting a seriers of beeps gettting closer together the closer you came to the phone May 04 12:24:25 counter May 04 12:24:25 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, 3 days 11:35:32 (10.483 +-3.5 days) (1087;194) May 04 12:24:39 counter? May 04 12:24:40 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) source: http://wiki.openmoko.org/wiki/MokoMakefile ; P0: http://wiki.openmoko.org/wiki/Wishlist:Neo1973_P0_Review; a week, 3 days 11:35:18 (10.483 +-3.5 days) for devices for *ANYBODY* for $350 (targeting developers) (preorders: not yet) (2007-05-13); 4 months, a week, 4 days (134.983 +-15.0 days) for mass market (2007-09-15): see topic for more info (1088;194) May 04 12:25:06 i want to preorder mine... May 04 12:30:23 abraxa_ - AFAIK there will be a lanyard. Not that I want one. I would rather opt for a screen protector thingy. May 04 12:31:12 Elrond: A screen protector of what kind? May 04 12:32:29 abraxa_ - some foil. May 04 12:33:22 Oh, those... a friend of mine attached one to his Treo display - it kind of "sticks" to the stylus so sliding across the screen doesn't work anymore May 04 12:33:46 I wonder if that's the same with all such screen protectors May 04 12:34:14 I don't know. I'd hope no. May 04 12:34:41 hmm.. binary gps module within 2 weeks? -> shipping of P1 withing 2 weeks? May 04 12:34:57 abraxa: no, most protectors are actually ok. May 04 12:35:11 * aloril guesses they will try to get working gps (including closed binary module) in P1 May 04 12:35:50 Two weeks does sound like a sound estimate May 04 12:35:57 aloril - They should. They don't want all people to send their P1 back, because the gps is noticed later to be borked. May 04 12:36:33 Elrond: they could test it works and still send without binary module, but I guess that would be bad PR May 04 12:37:33 scroll < 3h upwards for relevant quotes May 04 12:38:10 well.. s:module:library: May 04 12:39:02 I wonder why they used a soft-gps when it was foreseeable that no open source driver would be supplied by GL May 04 12:39:36 I'm guessing price May 04 12:39:58 Probably... May 04 12:40:06 And maybe space too, I don't know how big the hw parts are May 04 12:41:24 am i right that to use the gps information will need to be sent to the network provider? May 04 12:41:53 webjames: no. May 04 12:42:07 webjames - Read the FAQ for the GPS in the wiki. May 04 12:42:19 pavelm: cool, i did i was just checking i understood it May 04 12:42:53 pavelm - You decoded some part of the protocol? Did you put the info in the wiki already? May 04 12:43:40 * aloril thinks that eventually it might turn out to be better solution once GPL driver has been written May 04 12:45:20 aloril - Well, we still have no lawyer answer on this whole issue. May 04 12:48:27 Elrond: lawyer answer? May 04 12:49:34 Elrond: hammerhead protocol. Not much is done, but i suspect cmd 3 messages are gps data, and someone else deciphered how to tell it to tune to particular sattelite. May 04 12:49:54 * aloril remembers export problems with crypto, still crypto software was written ;-) May 04 12:49:54 pavelm - There are two issues: a) The old simple "reverse engeenering" issue (clean room, etc. pp) May 04 12:50:52 pavelm - b) GPS without some restrictions (not working at large speeds / altitudes) are considered munition in the US. May 04 12:51:17 i thought you had the right to bear arms there May 04 12:51:17 SCNR May 04 12:51:21 and a GPL driver has a self documenting switch for those trestrictions. May 04 12:51:42 buz - export/import is the issue, not havig one inside the US. May 04 12:51:54 elrond: a) we don't need clean room as we are not reading they disassembly but traces of communications and b) we simply make it limited. I do not think source == documentation May 04 12:52:01 well the main issue is that you can travel from the us with one May 04 12:52:11 buz: an unrestriced GPS is mentioned in the next paragraph of to the law on importing/exporting nuclear weapons test equipment. May 04 12:52:18 elrond: it is just a source... it is only ammunition if you compile it with switch removed and load it on the hw. Don't do it then. May 04 12:52:30 pavelm: Are you however a lwayer with ITAR experience? May 04 12:52:48 speed: I live in free country :-) May 04 12:52:58 The penalties are potentially unlimited fines, and long jail terms. (for FIC people). May 04 12:53:28 fic can continue to ship the closed source driver May 04 12:53:34 I'm not saying it is an issue. Just that it is not obviously not an issue. May 04 12:53:35 yeah. May 04 12:53:36 when crypto was under same laws, even USA citizen wrote one (PGP) ;-) May 04 12:53:37 ...for anyone caught travelling with limitations-removed-from-gpsd... May 04 12:53:41 BTW: Doesn't the same problem exist basicly for tomtom? May 04 12:53:44 obviously they cant be held responsible for their users actions May 04 12:53:47 Elrond: No. May 04 12:54:09 Elrond: Essentially all GPSs ship with updatable firmware and binary blobs for the GPS code. May 04 12:54:13 and that is with it enabled, not with "default off", so kind of worse May 04 12:54:17 ...plus it would probably be trivial to remove 1000kph limit from binary driver. May 04 12:54:35 However, it's trivial for _all_ the flashable devices. May 04 12:54:42 Which is the majority nowadays. May 04 12:54:59 SpeedEvil - I mean: The straces are from tomtom, so the "to be written gpl hammerhead gpsd" might work there too. May 04 12:55:02 if FIC people are scared about going to jail... they should just place sirfIII there. May 04 12:55:16 A binary blob would be perfectly safe. May 04 12:55:28 Not to mention that the people who *want* to build such ammunition will do so regardless of those laws May 04 12:55:29 And sirfIII is significantly poorer performing. May 04 12:55:35 source code is perfectly safe for FIC people, too. May 04 12:55:54 I mean -- they have nothing to do with it. May 04 12:56:08 ...and probably can't help us because NDAs, anyway. May 04 12:56:11 If they ship it - not. May 04 12:56:17 abraxa_ - Right. Not to mention. That doesn't stop a judge from sending people to jail. May 04 12:56:31 If it's a user upgrade, they can point to the 'evil hackers' doing it. May 04 12:56:46 Much as any other GPS maker in the same position having users replace the firmware could. May 04 12:57:12 Elrond: Nothing stops a judge from doing ridiculous things in the US indeed :) May 04 12:57:33 speed: even if they shipped binary gpsd (derived from GPLed code0 with 1000kph limitation... I do not see how they could get into jail for that. ammunition limit is for "whole devices". May 04 12:57:36 And you don't even have to be in the US, or be a US citezen. May 04 12:57:37 abraxa_ - Well, in this case they have a law they can point at. ;) May 04 12:58:05 Software + device is reasonably the "whole device". May 04 12:58:13 speed: US laws do not apply to _me_. I do not care about them. If FIC people are not in US, they are not bound to them, either. May 04 12:58:21 pavelm: Wrong. May 04 12:58:35 There are a number of countries with which the US has extradition treaties. May 04 12:58:41 Google 'natwest three'. May 04 12:58:48 what about if they sold it as hardware only? May 04 12:58:59 SpeedEvil: and Mckinin May 04 12:59:06 SpeedEvil: or however you spell his name May 04 12:59:06 speed: Yes, so if you ship software + device from US, you should make sure to have 1000kph limit there. May 04 12:59:57 The problem as I see it is, device + software is clearly a 'GPS' within the meaning of the law. Does the fact that it's got a documented 'off' switch to the restrictions make it unrestriced in the eyes of the law. May 04 13:00:22 Into or out of US. May 04 13:00:31 speed: Anyway, it is FIC who will have to consult lawyers. Not us. May 04 13:00:39 Yeah. That's all I've been saying. May 04 13:01:25 The hammerhead is fundamentally more functional than the sirfIII, which is really good. May 04 13:01:34 Especially if we can get a GPS driver for it. May 04 13:01:38 s/S/L/ May 04 13:01:38 SpeedEvil meant: Especially if we can get a GPL driver for it. May 04 13:01:43 hmm, someone should invent binary patch for disabling 1000kph limit from tomtom... and have some fun. May 04 13:01:54 That's not tomtoms problem. May 04 13:02:04 Anyone around who knows the ways of the OpenMoko OE overlay? May 04 13:02:05 They simply point to the hacker, and say it's their fault. May 04 13:02:27 Yes, so if someone points out how to remove if (speed >1000) from the GPLed code, point to the hacker and say... May 04 13:02:51 and hammerhead is piece of crap and will stay so for at least next year :-(. May 04 13:03:08 The hardware is cool - if it came with docs. :/ May 04 13:03:13 SpeedEvil - Are there chances for hackers (like us) to get problems, when we create the "off-switch" (when we write the GPL driver, we create the off-switch). May 04 13:03:23 Elrond: No. May 04 13:03:36 Elrond: the law explicitly says whole GPS units. May 04 13:03:41 GPS software is not covered. May 04 13:04:10 However, the problem isn't it getting confiscated at customs from an individual and tested. May 04 13:04:34 It's apple/... pointing a tame prosecutor at the importer. May 04 13:04:38 Couldn't it be considered "helping to create munition"? May 04 13:04:41 (of many units) May 04 13:04:51 Elrond: I don't think so. May 04 13:05:09 speed: Yep, the same way winmodems are cool. They are "fundamentally more capable", same way hammerhead is, but they are also pain :-( May 04 13:05:10 The offence is simply importing a GPS capable of reporting at high speeds. May 04 13:05:23 It's not assembling one. May 04 13:05:28 importing/exporting May 04 13:05:31 why is reporting at high speeds an offence? May 04 13:05:44 webjames: Think guided missiles May 04 13:05:47 Basically as they could be used as ICBM guidance. May 04 13:05:49 Anyway, I should go for some outside sport. May 04 13:05:51 webjames - could be used for rocket ;) May 04 13:06:02 looking at some devices they report upto different speeds May 04 13:06:04 The law was written back when making a GPS was much harder. May 04 13:06:27 webjames: If you import *anything* to US at >1000kph, I guess it is called terorism :-). Remember that Concordes are grounded. May 04 13:06:30 So it was slightly less silly. May 04 13:06:48 webjames: ...and us probably will not like you firing missiles at them. May 04 13:06:49 pavel ;) May 04 13:07:04 webjames: ...er... importing missiles at high speed :-) May 04 13:07:07 yes, i see now the problem May 04 13:07:40 "importing missiles at high speed" sounds fun. ;) May 04 13:07:43 Ugh... anyone knows what's up with openmoko-today? May 04 13:08:12 Elrond: don't say that aloud, they may charge you with terrorism ;-). May 04 13:08:31 pavelm - Yeah. May 04 13:08:34 Elrond: actually I thought that sending model aircraft to US with payload of sugar may be a lot of fun ;-). May 04 13:08:55 pavelm: Jef Raskin had the same idea... He's Dead now. :) May 04 13:08:56 ...(and to stay ontopic, we could use openmoko for navigation) May 04 13:09:13 Also, the $1K cruise missile guy got into all sorts of tax problems. May 04 13:09:15 my uncle was telling me (i don't believe him) that in New Zealand all th gps's are made inaccurate May 04 13:09:18 200 years ago, it was inquisition. May 04 13:09:27 webjames: Google selective availability. May 04 13:09:41 webjames: GPS used to be made intentionally inaccurate. May 04 13:09:58 What about a ICBM that only goes 999km/h? :-) May 04 13:10:08 SpeedEvil: but not any more May 04 13:10:10 50 years ago, it was "communism" May 04 13:10:11 Okay, bye for now. May 04 13:10:24 Today it's ... fill in the right word. ;o) May 04 13:10:32 pavelm :' wave May 04 13:10:40 stupidism? May 04 13:10:40 pavelm - Bye :) May 04 13:11:20 SpeedEvil - Yeah. ;) May 04 13:11:42 I wonder when the first GPS/Galileo combi modules will be on the market which make sure that there won't be any intentional misguidances May 04 13:12:23 GPS/GLONASS are already available. May 04 13:12:31 GLONASS? May 04 13:12:34 Never heard of that before May 04 13:12:51 I'm not sure if any of the chipsets support the more precise GLONASS version that was made available for civil use earlier this year. May 04 13:12:54 The russian system. May 04 13:12:59 glonass is the russian system. May 04 13:13:03 It was in disrepair till a couple of years ago. May 04 13:13:10 With few satellites. May 04 13:13:14 More are being launched. May 04 13:13:24 There is already one or two modules that do GPS and "galileo ready". May 04 13:13:24 India is involved in funding AIUI. May 04 13:13:37 Interesting... is China going to orbit a navigation system, too? May 04 13:13:49 Who knows. May 04 13:13:56 abraxa_ - I heard, that china plays in the galileo game. May 04 13:14:24 Ah... good... it's quite odd to have 3 navigation systems up there as it is May 04 13:14:45 Usually I think, it's good. May 04 13:14:56 More redundancy. May 04 13:15:03 Ideally. May 04 13:15:16 You'd want one system with multiple satellites. May 04 13:15:23 True, except that I doubt they are put up there for civil/research reasons May 04 13:15:28 I mean multiple over what's required in clear horizon. May 04 13:16:01 abraxa_ - galileo was intended for civil use, no? May 04 13:16:07 abraxa_, May 04 13:16:19 SpeedEvil - redundancy in "organisations" controlling the system. ;-) May 04 13:16:24 That too. May 04 13:16:28 Do you really believe what the government tells you? ;) May 04 13:16:33 Sup3rkiddo: Sup Superkiddo May 04 13:16:41 abraxa_, any luck mate.. May 04 13:17:46 With what? Adding a package? May 04 13:17:51 abraxa_ - well, the gps use was clear I think, from nearly the start. So they don't lie always. May 04 13:18:08 abraxa_, yup.. May 04 13:18:21 maybe we should launch OpenPositioning? May 04 13:18:51 Sup3rkiddo: Yeah, I got it to work and will put it on the wiki once I have that stupid openmoko-today issue sorted out here May 04 13:19:27 abraxa_, sure would be great...thanks May 04 13:19:50 yupyup May 04 13:20:29 Elrond: I have long lost faith in governments, sorry :) May 04 13:20:36 If you can assume IP connectivity of the devices that get positions, tehn teh satellites get much easier to make. May 04 13:21:19 As you don't need quite so accurate clocks, and they just need to broadcast relative not absolute time. May 04 13:21:46 Without even needing to send orbit info, which lowers power. May 04 13:22:33 SpeedEvil - Hmm, I don't know, if I like that. May 04 13:24:28 abraxa_ - Well, I'm just sceptical and research things, if they really matter to me. Otherwise I consider the info as "might be wrong, use with care". May 04 13:24:28 (script) wiki RecentChanges: [[Wish_List_-_Hardware]] [[Wishlist:Text_Input]] May 04 13:24:30 Yeah - being independant of the net is a big plus. May 04 13:43:07 hi May 04 13:43:58 hello guillaum1 May 04 13:44:26 hi arw May 04 13:46:10 Does someone know if an audio server is planned, like pulseaudio or esound ? May 04 13:46:30 pulseaudio rather May 04 13:48:35 hrw, the TCP module of pulseaudio could be avalaible ? May 04 13:49:18 or UDP ? May 04 13:51:13 http://www.openembedded.org/filebrowser/org.openembedded.dev/packages/pulseaudio/ - read pulseaudio_0.9.5.bb May 04 13:53:14 hrw, cool ! thanks for the link May 04 14:09:00 aloril: counter May 04 14:09:01 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, 3 days 09:50:57 (10.410 +-3.5 days) (1089;194) May 04 14:16:02 i think this whole gps thing is ridiculous, i'm going to ask a lawer what he thinks might be the verdict May 04 14:16:47 i think the eurasian fractions doesn't care and the us people should tidy up their jurisdictal system anyways :) May 04 14:19:24 qd33333* May 04 14:20:48 cleaning up the keyboard ;) May 04 14:21:28 webjames ? May 04 14:21:54 roh ? May 04 14:22:03 webjames i do not see a real problem there at the moment (besides licensing/technical stuff) May 04 14:23:15 roh: it's the problem with the US ruling a gps reading above 1000knots and 60 000feet ruling a said gps device as a munition May 04 14:24:10 webjames so what? as long as we do not mount jet-engines on the moko it will not really happen May 04 14:24:49 no, but it could, and thats the problem.. a bit annoying considering i live in the UK May 04 14:24:57 a device which doesn't disable itself when being faster or higher than that is munition May 04 14:25:25 roh: what he said May 04 14:25:35 webjames sorry.. airstrip one ;) May 04 14:26:31 Hopscotch then we build in some artificial software limit. what do i care if a customer patches his software (could do that with every other gps module also) May 04 14:27:02 roh: doesn't need to - i had that limitation disabled in a gps device here myself May 04 14:27:19 you just send it back to the manufacturer and tell them, that you're not american May 04 14:27:41 roh: that's exactly my point, if a user hacks the software it's not the manufactures fault. unless it was ruled that they didn't try hard enough to make it non-hackable May 04 14:28:06 is any american here anyways? May 04 14:28:23 england here May 04 14:28:44 germany here May 04 14:28:47 webjames i do not know any law which has definitions of 'hard enough to hack' (which is silly anyways) May 04 14:29:05 i guess most ppl here are european anyways - so just keep on hacking :) May 04 14:30:07 roh: i agree, it's just live DVD regions, i suppose such a thing could be implemented into a gps, but the you would get 'region free' devices which any american could import May 04 14:30:54 Hopscotch: yep May 04 14:31:21 basically i think it is a stupid law, as if someone wanted to disable the limit they could with a simple hack May 04 14:32:03 You guys miss the point. Obviously if a GPS travelling at 1000 knots at 60,000 feet hit you, it'd hurt a lot. That's a munition! May 04 14:32:06 * dho ducks May 04 14:32:39 dho - nice, but then it's only accurate to 5meters or so. i'll take my chances - i say disable the limit! May 04 14:33:07 mickey|bbl: ping May 04 14:33:52 i think FIC will seek legal advice, but come to a similar conclusion to what we have come to. May 04 14:34:55 * dho doesn't see what you need a GPS operating for at 1000 knots / 60000 feet May 04 14:35:20 Or any combination of the two May 04 14:35:44 Rather, either of them separately :( My brain hasn't quite woken up yet. May 04 14:36:28 well, we wouldn't have any problem if the darn h/w manufacturers spent some money to integrate said limitations into the hardware itself May 04 14:37:49 CoreDump|home yes, but then it would be unfair on those who live outside the US May 04 14:38:14 why? May 04 14:38:35 What people outside the US have the capacity to operate the thing at 1200mph or 12 miles in the sky? May 04 14:38:45 What people inside the US have that capacity? May 04 14:38:48 they could ship two versions. But IMO the limits is far beyond any civilian application May 04 14:38:54 Precisely May 04 14:38:59 what if i want to track my journey whilst in an air plane? May 04 14:39:18 What airplane are you in that meets either of those two criteria? May 04 14:39:39 commercial jetliners travel at <1000 knots and < 60k feet May 04 14:39:46 the anti-gravity plane May 04 14:39:51 =) May 04 14:43:07 so, I take it nobody succeeded in hacking neos GPS based on the strace yet? May 04 14:43:24 nope May 04 14:43:37 to bad May 04 14:43:45 * CoreDump|home would really like to play w/ it May 04 14:43:46 we have first clues how the protocol works, but the gps didn't give any statement to that yet May 04 14:43:55 "00" heh May 04 14:46:19 so far we're not sure if the gps in the neo is powered up anyways May 04 14:46:27 * Hopscotch is waiting for p1 May 04 14:48:19 if you have a tomtom one v2, you can test with that May 04 14:48:25 it uses the same gps May 04 14:48:48 what will be the status of this in a couple of weeks time, surly with unresolved issues and potential problems it cannot be released, exported or imported May 04 14:50:47 we're not depending no the gps-reversing effort May 04 14:50:58 global locate will provide a nonfree package for the gps May 04 14:51:02 gps will work anyways May 04 14:55:16 i see, there's no way global locate will be releasing it under GPL? May 04 14:55:48 nope, they're very proud of their great work... May 04 14:55:49 not that i can see May 04 14:56:14 atm we will have to wait and see what happens with the GL solution May 04 14:56:45 the licensing issues are not yet worked out May 04 14:56:59 i have yet to measure how much cpu power is needed for the gl app May 04 14:57:07 on the tomtom (which uses the same cpu) May 04 14:58:01 what kind of software could be expected to utilise the gps? would it be the openstreetmap, or a non-free solution? May 04 14:58:32 webjames i think we are not fixed yet there. May 04 14:58:50 what people create May 04 14:58:55 and how the project continues May 04 14:59:07 but atm we strongly prefer open software of any kind. we will have to see if that allows using maps which need to be paid May 04 14:59:07 webjames: the nmea stream will be available May 04 14:59:24 if a commercial provider likes the platform, i see no reason they wouldn't port their apps to openmoko and linux May 04 15:00:51 yeah, it do prefer opensource, however i do understand that mapping takes a huge amount of time and effort May 04 15:02:16 you saw the gaia project? May 04 15:04:04 roh - Do you know, wether FIC's lawyers are investigating the legal issues nehind a GPL driver for the GL chip? May 04 15:05:19 good question - but i don't think so May 04 15:05:38 Elrond there are discussions, but i cannot give you any informations on the current state May 04 15:05:44 since gl wasn't open to gpl anyways May 04 15:06:10 roh - They're aware of the ITAR regulations issue and trying to investigate this? May 04 15:06:29 i think so May 04 15:06:41 thats not a major problem May 04 15:06:51 It is. May 04 15:07:30 Anyway, I need to run. Back in a few hours. May 04 15:08:16 roh: ah, i should look where ppl are speaking :) seems you know much more than me :) May 04 15:08:27 where are the units being exported from? china? May 04 15:08:47 Hopscotch afaik yes May 04 15:08:57 eh webjames i meant May 04 15:09:09 Hopscotch i should, since i work for them ;) May 04 15:09:37 roh: :) /me has had some talks with harald but then the status was that gpl is not an option May 04 15:09:41 so the issue with the munitions should only effect certain countries. May 04 15:09:43 glad to hear that that changed May 04 15:09:57 Hopscotch afaik that has not changed. May 04 15:10:38 roh: :( May 04 15:10:44 Hopscotch its not only a driver. its their whole ip in that blob, since the GL chip does not have its own cpu (mostly arm7/9) like the ublox or sirf always have) which makes it much cheaper May 04 15:11:35 roh: yeah, is saw that. the gl is mostly an analog part with integrated serial port - no intelligence in there May 04 15:11:40 but we are taking care of it. there will be a acceptable solution for all of us. and since the coreteam (harald in front) is totally OSS fanatic you do not need to worry May 04 15:12:18 roh: i'm glad about that May 04 15:12:34 roh: i know - but then the fsf would recommend the phone if it would be totally open May 04 15:12:59 roh: and rms blessing would be something :) May 04 15:13:42 Hopscotch the fsf will never get a phone buiÃlt, then. gsm e.g. will never be open since it cannot get a certification then May 04 15:14:26 roh: afair rms was mostly concerned about the software running on the ap May 04 15:14:53 roh: that periphery has it's own cores running something else locally seemed to be ok for him May 04 15:14:59 Hopscotch same here. but be shure that we will do what we can to avoid that May 04 15:15:27 and we will support that from the other end of options :) May 04 15:22:43 other than the GPS there's nothing else which is unresolved to this kind of extent is there? May 04 15:23:00 there are plenty of unresolved issues May 04 15:23:11 and moreso uncertain May 04 15:23:49 o :( May 04 15:24:05 i ran an emulation with qemu, and i thought it was all looking really good May 04 15:36:09 Hadaka: are you naked on irc? :-P May 04 15:39:07 aevin: considering the nick, I'd say jime May 04 15:44:32 hi May 04 15:46:11 hewwo May 04 15:57:33 Anyone here who knows why openmoko-today is referenced in task-openmoko.bb but not present in OpenMoko's OE overlay? May 04 16:01:46 no idea May 04 16:03:14 :| May 04 16:05:05 gtg, cu May 04 16:39:18 balrog-kun: You around? May 04 17:00:16 What store's are going to carry openmoko? May 04 17:00:17 xamox: See http://wiki.openmoko.org/wiki/FAQ (Hopefully this link answers your question. This autotrigger for new nick and question during first 15 minutes done only once.) May 04 17:00:29 Undetermined basically. May 04 17:01:12 I wouldn't count on any of the big carriers picking it up May 04 17:03:10 It's GSM based correct? May 04 17:03:15 yes May 04 17:03:31 So in theory I'll be able to yank out my little card and toss it in there May 04 17:03:36 and have service? May 04 17:03:36 Yep. May 04 17:03:45 In practice, too ;) May 04 17:03:47 Unless your carrier whitelists phones or something. May 04 17:03:52 Most don't. May 04 17:04:08 Alright May 04 17:04:40 How much storage space is it suppose to have? or sd/cf card or something? May 04 17:05:09 microsd May 04 17:05:18 awesome May 04 17:07:06 xamox: http://wiki.openmoko.org/wiki/Category:Neo1973_Hardware has all that and more ;) May 04 17:08:03 Reminds me a lot of my sharp zaurus May 04 17:39:19 abraxa_: yeah May 04 17:44:08 balrog-kun: Did you already make it so usb-gadget support is optional? May 04 17:46:02 abraxa_: yep, ./configure tries to compile a short snippet that uses usb-gadget and if that fails it disabled CONFIG_GADGETFS May 04 17:47:02 Hmm May 04 17:47:23 disables* May 04 17:48:30 Any idea what that could be about then? http://paste2.org/p/2988 May 04 17:49:34 abraxa_: is this on a linux host? May 04 17:49:38 yup May 04 17:50:05 2.4 maybe? May 04 17:50:09 it's missing linux/usb_ch9.h and other fundamental stuff May 04 17:50:35 linux/errno.h May 04 17:55:35 I remember you mention that /usr/include/linux isn't always en par with the installed kernel May 04 17:55:51 So I copied /usr/src/linux/include/linux/* there May 04 17:56:00 Was this a bad move? May 04 17:57:10 abraxa_: no, should fix all of these errors May 04 17:57:49 but with all these files missing i think most programs wouldn't build May 04 17:58:18 Well, I seem to have missed some files and just added them, resulting in this: http://paste2.org/p/2989 May 04 17:59:26 Does qemu require a certain kernel version by chance? I'm running on 2.6.20.3 May 04 18:00:10 i think it should run even on 2.4 May 04 18:00:28 basically on Unix clones in general May 04 18:01:45 it even runs on osx May 04 18:02:02 Hmm. May 04 18:02:07 balrog-kun: any ETA on getting the neo1973 stuff upstream? May 04 18:02:19 abraxa_: does a Hello, World! using compile on your system? May 04 18:02:54 koen: hmm, hard to say May 04 18:03:05 probably not longer than two months :P May 04 18:03:10 ah, cool May 04 18:03:11 koen: is there any reason to hurry? May 04 18:03:31 no, just curious May 04 18:03:42 seeing that you got the zaurus stuff in this week May 04 18:04:16 there is a conflict between this two implementations of s3c2410 May 04 18:04:35 and there are two more on-chip peripherals that iw ould like to add before upstreaming May 04 18:04:42 out of like 22 May 04 18:05:11 thanks for the heads up May 04 18:05:31 :) May 04 18:06:59 balrog-kun: Yeah it does - I wiped the build/qemu dir completely and restarted the build, got the same conflicts again May 04 18:07:26 Should I do something with the /usr/include/sys dir maybe? May 04 18:08:12 usr/include/linu shouldn'T include the headers of a the kernel source at all May 04 18:08:30 right May 04 18:08:32 What else then? May 04 18:08:36 userspace applications are supposed to build against linux-kernel-headers a seprate package May 04 18:08:46 abraxa_: i'm not sure, never had this exact problem May 04 18:09:26 balrog-kun: Alright, thanks - I'll fiddle around some more then :) May 04 18:10:01 abraxa_: okay, i hope it works out May 04 18:10:27 it looks like two different header have defined the same identifiers May 04 18:10:32 headers May 04 18:10:50 Yeah, though how that happend I have no idea May 04 18:12:20 your /usr/include/linux contents is identical with /usr/src/linux/include/linux contents/ May 04 18:12:32 Yup May 04 18:12:33 s/\/$/?/ May 04 18:13:06 counter May 04 18:13:07 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, 3 days 05:46:51 (10.241 +-3.5 days) (1090;194) May 04 18:13:19 Fortunately it's just a chrooted sandbox so it's not critical for my system - it does however prevent me from building qemu so it needs fixing :) May 04 18:30:09 consider these 2 cases: May 04 18:30:11 1) source code for existing closed GSM module is leaked and also way to update firmware with modified one May 04 18:30:11 would this really make phones based on this module illegal? May 04 18:30:11 2) s/leak/GPL/ May 04 18:30:11 May 04 18:30:31 i've just built the latest and just about to run qemu now if i can be any help? May 04 18:30:52 compiled firmware with hardware is certified, not source code May 04 18:30:57 leak it May 04 18:31:13 :-) May 04 18:32:40 add some disclaimer like this: "You are allowed to change, like GPL says, but then *you* are responsible for getting recertication from FCC" May 04 18:32:59 consider: any organization that has the resources to build and fire an ICBM also has the resources to build a guidance system for such rocket May 04 18:33:49 aloril2, the comparison makes a good point, but try making it to the lawyers... May 04 18:34:29 lawyers are the scum of the earth--praying on the people's misfortune May 04 18:34:40 prey? May 04 18:34:43 Except the ones that can spell. May 04 18:34:47 example: adding crypto at lower level: needs certification May 04 18:34:50 yuk May 04 18:35:08 Speed: i got there first May 04 18:35:24 :-) May 04 18:36:13 what do you have with 12 lawyers buried up to their necks in sand? May 04 18:36:40 not enough sand... May 04 18:37:53 I think we should fly under the radar--develop our own GPS spftware and let them try to stop us May 04 18:38:02 software May 04 18:40:04 DukeOfURL, also not enough lawyers... May 04 18:40:21 A really open GSM would require a softradio, since no existing GSM radio will opemn up the specs for their baseband May 04 18:40:49 DukeOfURL, since the desire to build a free GPS driver is rather public as is preliminary protocol analysis, I don't think under the radar is much of an option :] May 04 18:40:53 Since we started with broken analogies: Guns can be used to commit crimes, still you can sell them since it's the user's responsibility. Configurable radios can be used to violate FCC rules... why can't this be user's responsibility? May 04 18:41:27 because the spectrum is deemed more important than a life of a human being apparently May 04 18:41:29 You said GPS, not GSM. I though the conversation started with GSM? May 04 18:41:36 ...and the obvious reply is that someone misbehaving with a gun is easier to track than a zillion elbonian radios that are out of spec. May 04 18:41:44 Or are you talking about location services? May 04 18:42:02 ...and collectively mess up the spectrum commons for everyone May 04 18:42:19 GPS - the restrictions placed on the GPS receiver at extreme speed and altitude May 04 18:42:25 Ahhh May 04 18:42:58 Yeah, that will be a tough problem without truly open firmware May 04 18:43:20 EIther that or the discovery of how that is configured in an existing device May 04 18:44:35 * Paavo_ is wondering how big of a deal would cooking up a homebrew GPS be for The Terrorists(tm)... May 04 18:45:43 let's go back to the original arguments concerning hacking HBO--these signals are arriving onto my property, and I can do whatever I want with such signals May 04 18:47:14 HBO: Home Box Office, a US movie channel May 04 18:51:27 Terrorists don't need that kind of GPS. Those restrictions are more aimed at North Korea, so they can't build accurate missiles. Though it is irritating for amateur rocketry. May 04 18:52:58 Tell me about it. Can't even order H2O2 here in that states any longer May 04 18:57:12 Zigurd, the GPS problem doesn't require open firmware. The GPS chip in the Neo is quite dumb, the position calculation and the US statutory limitations are in the userspace driver, which is presumably possible to reverse-engineer. May 04 18:57:33 make installing FCC uncertified firmware enough cumbersome so that only developers do it (== test equipment then, like software radios are sold) May 04 18:57:45 does anyone have a neo1973? May 04 18:57:59 seemant: Yes, about 40 people May 04 18:58:17 abraxa_: so those 40 are working on apps and stuff? May 04 18:58:32 Yes, that's why they have them May 04 18:58:37 or can anyone with "tree fitty" get one? May 04 18:58:44 Not yet May 04 18:59:07 and that projects url is where all those apps are then? May 04 18:59:16 What is the GPS strategy in the Neo? Is the GPS completely independent? Or is it like AGPS? May 04 18:59:21 the "projects" link on openmoko.org I mean May 04 18:59:47 it the interface slow just on qemu? May 04 18:59:49 Zigurd, you're confused. It can work either assistedly or independently. May 04 18:59:58 but I guess GPL GSM module will be based on software radio, not on some specific chip(s) May 04 19:00:04 seemant: wiki.openmoko.org is more like it - projects.openmoko.org hosts all things not done by the core team May 04 19:00:40 ahh I see May 04 19:00:53 aloriÃl2 there will not be a gpl sm anytime soon as far as i can see it May 04 19:01:20 mjr: so the gps thingy -- we'll be able to have maps on the microsd card and use the phone as a navigational device? May 04 19:01:29 i've just compiled 'make setup' and then 'make qemu' everything looks like it's coming along great but the interface is quite slow, will this be the case just on qemu? May 04 19:01:36 like a garmin or something May 04 19:01:40 abraxa_: thanks! May 04 19:01:53 seemant: Welcome :) May 04 19:01:58 seemant, yes, though free maps are... bad quality May 04 19:02:18 mjr: how about a google map mashup type web2.0 buzzwordy thingy? May 04 19:02:45 and proprietary ones aren't well available unless somebody starts selling their nav software for moko May 04 19:03:18 google map can be interfaced to, its license (which may apply or not according to where you live) doesn't permit caching though May 04 19:04:08 For just playing around, you can get free access to Autodesk, if you ask nicely. May 04 19:04:13 (I'm pretty sure many will disregard that, and I would too, if my city didn't provide good local maps, bitmap only though) May 04 19:04:41 But for most apps, Google is good enough May 04 19:07:56 roh: I think GPL software radio GSM will happen without asking 'permission', like PGP happened May 04 19:11:18 aloril2 since it will take very long till it 'works' and 'would get through certification in compiled form' these radios will cause so much hassle they will take up the work to triangulate you May 04 19:11:51 There are GSM softradios, but they were multi-million dollar commercial developments, like Vanu. May 04 19:12:19 It will be many years yet before you can talk on a GSM radio using a really open source device and software. May 04 19:12:59 Might as well do VoIP on WiMAX instead, if you want a secure, private means of communication. May 04 19:15:29 yea gsm isn't the end-all... May 04 19:16:22 "GnuSM" :) May 04 19:17:12 can the usrp do gsm? May 04 19:18:27 Unfortunately, it'll probably at least a decade before we see the likes of a Ettus Gnuradio on a 10mW chip... :/ May 04 19:19:42 koen not in realtime... means there is no single pc i know of which can do gsm completely in software May 04 19:20:53 balrog-kun: Is there any way I can manually disable usb-gadgets support? May 04 19:32:04 abraxa_: in config-host.h undefine CONFIG_GADGETFS May 04 19:32:39 config-host.h is created durig ./configure run May 04 19:32:45 during* May 04 19:35:52 balrog-kun: Ahh, thanks... now it finally compiled :) May 04 19:43:59 roh - Why do you think, that the ITAR issue is trivial? May 04 19:47:43 Elrond because as far as i can see it we will not get any working open software for gl soon, so there is none. else i do not know what we will use in future products May 04 19:50:17 roh - Well, people are already working on decoding the protocol. And after getting the pseduo-ranges or something alike, the rest is only plain math. May 04 19:50:57 roh - And those working on it, are concerned about the possible issues. May 04 19:56:03 just add some software limit at the said parameters. then nobody can say you violate smth. that stupid. when flying 1000mph you usually have other problems May 04 20:49:33 i've just compiled 'make setup' and then 'make qemu' everything looks like it's coming along great but the interface is quite slow, will this be the case just on qemu? May 04 20:52:43 Yup May 04 20:52:58 sweet May 04 20:53:05 You're basically emulating completely different hardware May 04 20:53:27 oh right, awesome.. thats good May 04 20:53:33 abraxa_: ? May 04 20:53:43 arm is a heavy different instruction set from x85. May 04 20:53:43 abraxa_: it emulates the same hardware May 04 20:53:49 s/85/86 May 04 20:54:42 webjames: it depends on your computer's speed but afaik on my PC (2.4 GHz) the UI is faster in qemu than on Neo May 04 20:54:45 but the emulation isn't accurate? May 04 20:55:24 why it has to? May 04 20:56:05 my cpu is 3.4ghz... does this mean i'm experiencing accurate UI speeds? May 04 20:56:43 no. it means that intel engineers forgot that GHz does not mean speed May 04 20:57:28 yeah, also it's HT which is crap with linux, it picks up two cores so it's really only using half i guess.. big love to AMD when i upgrade May 04 20:57:57 Why would HT be crap with linux? May 04 20:58:26 because of the way it sees the chip, it splits it in two, but really it's one so can't act as two May 04 20:58:53 webjames: you use 2.4 kernel - right? May 04 20:59:28 hrw: 2.6.20-15-generic May 04 21:00:42 i do like the stroke recogniser! May 04 21:01:07 webjames: read the [[OpenMoko_under_QEMU]] page about UI speed - no emulator will give you accurate speed May 04 21:01:52 i did read that, but then i also heard that it was faster, so i was wondering what was what May 04 21:02:10 yeah, for most stuff it's faster May 04 21:02:31 i mean it's not slow slow, but takes about 6seconds to load up the contacts May 04 21:02:55 on the phone it probably takes more than that, but i don't know May 04 21:03:14 really? .. May 04 21:03:16 i'm quite sure it will be optimised before P2 May 04 21:03:32 balrog-kun: you lack phone hardware? May 04 21:03:47 yeah cos it's almost 10 seconds to load the calender May 04 21:03:53 hrw: yep, i only have the virtual one May 04 21:04:33 it's ok but i can't phone my friends from it :( May 04 21:04:52 nice, :) May 04 21:04:56 hrw: I guess that's the reason why he started work on QEMU in the first place ;) May 04 21:05:24 not exactly, but that's a factor too :p May 04 21:06:32 abraxa_: wrong guess ;D May 04 21:08:06 i see phase 1+ has a faster CPU May 04 21:08:51 GTA02 phone will get s3c24xx probably May 04 21:10:15 also a discount for those who brought a phase 1 device, that's nice, although doesn't say how much? May 04 21:11:11 webjames, details promised when sales start May 04 21:11:50 mjr: yes so i see, a nice way of rewarding devs though :) May 04 21:12:32 faster hardware is not a remedy for slow programs May 04 21:12:36 it's a workaround May 04 21:13:08 well, for some things you just need the power May 04 21:13:13 cu May 04 21:13:43 and i believe OpenMoko *is* slow in current version, but that's because there was no pressure on the speed until now, but can probably be fixed later May 04 21:13:46 balrog-kun: i agree May 04 21:14:14 (but that's stuff like media decoding and the like) May 04 21:14:14 hrw- bye May 04 21:15:05 mjr: yeah video i suppose might be heavy on resources.. be good to get that running smooth May 04 21:15:52 wiki is promising though for 320x240 video playback on current hardware May 04 21:16:08 mjr: Which would be awesome. May 04 21:16:15 mjr: current as in phase1? May 04 21:16:35 does it have wifi ? May 04 21:16:35 P1.5 will likely have WiFi (Atheros AR6K): http://lists.openmoko.org/pipermail/announce/2007-April/000012.html http://wiki.openmoko.org/wiki/Neo1973 http://wiki.openmoko.org/wiki/FAQ#Q:_What_is_the_rationale_behind_the_exclusion_of_WiFi.3F (WiFi status for P1.5 will be known for sure when P1 becomes available and also what kind of discount for P1.5 is given for P1 buyers) May 04 21:16:43 webjames: phase0/phase1, yep May 04 21:16:45 webjames, yes. http://wiki.openmoko.org/wiki/Video_Player May 04 21:17:43 do you think i could stick a standard xvid on it and play it at phase1+? May 04 21:17:56 wouldn't count on it May 04 21:18:36 mjr: but with the res shrunk down to 320x240 @ 30fps? May 04 21:18:51 you'd still have to do software rotation and scaling (unless the res is okay as such) in addition to decoding May 04 21:20:03 yeah, i suppose video at 480x640 is out for the moment May 04 21:20:33 a lot depends on also the bitrate etc May 04 21:20:33 i was amazed to see the multitouch demos links from the wiki May 04 21:21:02 mjr:very true i'm only thinking of long train journeys.. May 04 21:21:11 but almost certainly decent neo-prepared video can be played on p1+ if it's almost there on even p1 now May 04 21:21:13 240*320 - the LCD can scale. May 04 21:21:29 So there is no software scaling problem. May 04 21:21:39 brilliant May 04 21:21:51 SpeedEvil, he was talking "standard xvid", I was assuming "any ol' xvid I get my hands on", which would likely involve scaling... May 04 21:21:58 Oh - right. May 04 21:21:58 i used to have a k800i with that res, the videos were 3gp, but looked pretty good May 04 21:22:03 In that case, probaably not. May 04 21:22:16 I have doubts that 320*240 is really needed on a 2.8" display. May 04 21:22:51 SpeedEvil, true, but can the LCD manage scaling to smaller resolutions? May 04 21:22:57 Not AFAIK. May 04 21:23:08 However, 160*120 could be software scaled up. May 04 21:23:08 I do know it can do the 320x240 thing, which is good for video, but others... May 04 21:23:13 the k800i only had 2 inches May 04 21:23:28 SpeedEvil, hmm, true... May 04 21:23:55 you just have to hold it closer to your face May 04 21:25:15 well, anyway, the wiki says p1 does 400kbps 320x240(prerotated) mpeg4 at 25fps without sound and 18fps with, which is certainly a good starting point. May 04 21:25:25 yeah May 04 21:25:32 maybe use a crappier codec like mpeg1 ... May 04 21:25:43 yeah, certainly as p1+ will have faster cpu May 04 21:25:53 looks very promising May 04 21:26:02 or just give up and go 12fps :] May 04 21:26:26 wouldn't be the end of the world, though 25 is to strive for May 04 21:26:29 (script) wiki RecentChanges: [[Category:Neo1973_Hardware]] May 04 21:26:36 mjr, more than that would just be greedy! :p May 04 21:26:42 Consider that on most screens 160*120 video is a little smaller than the neo screen. May 04 21:26:51 Elrond ? May 04 21:26:54 oops May 04 21:27:16 yeah, they've packaged a lot of pixels! May 04 21:28:09 yes, it's a good point that straight-up 2x software scaling from 160x120 would probably be sufficient for most uses May 04 21:28:22 and light enough on the cpu May 04 21:28:43 yes, unless it was subtitled, in which case the res would be most useful May 04 21:29:01 that would be one case where you want res, yes May 04 21:29:09 also, you want to look at the screen very closely ;) May 04 21:29:32 or you could get a bluetooth HUD? May 04 21:29:53 Naah. All you need is a headmount. May 04 21:30:03 Hold the neo at 15cm, with a lens over it. May 04 21:30:10 :) May 04 21:30:16 yeah nice May 04 21:30:17 I'd buy that ;P May 04 21:30:40 I'll probably have to fix something up to attach a bluetooth keypad to my wrist or something... May 04 21:30:58 hold neo in left hand, operate keyboard on wrist with right... May 04 21:31:02 > I'll do the timing May 04 21:31:04 > ...phase 0 software May 04 21:31:04 > ...contacts launched ready within 15 seconds, window appeared in 10 sec. May 04 21:31:04 > ...dates ready wihin 11 seconds, window appeared in 7 or so. May 04 21:31:04 > ...second launch of contacts takes "only" 12 sec. May 04 21:31:04 > ...menu in contacts takes about 1 second to appear. May 04 21:31:05 * pavelm is afraid we'll end up with symbian-like "one big app" to keep latencies May 04 21:31:07 +down. May 04 21:31:07 yeah, i did like the bluetooth credit card keyboard featured on the input page May 04 21:32:09 pavelm: thanks for that and that was using p0 hardware? May 04 21:32:54 webjames: yep. May 04 21:33:01 pavelm: Is that pinning the CPU, or is it reading from flash. May 04 21:33:14 pavelm: well things can only get better May 04 21:33:43 speed: I do not have easy way to measure that... but given that re-start of contacts is 12 sec (vs 15), I'd say cpu-bound. May 04 21:34:02 Yuk. May 04 21:34:05 Oh well. May 04 21:34:14 Given that probably no time has been taken to optimise. May 04 21:34:59 a nice benchmark for improvements to be made on. i'll take down the figures pavelm and see how they compare later May 04 21:35:08 that optimization seems necessary though... May 04 21:35:27 Yeah - under a second's needed. May 04 21:36:21 * abraxa_ wonders if precaching binaries would help May 04 21:36:22 yeah, after using a desktop computer users are not used to waiting May 04 21:37:50 abraxa_:i would think it would well the popular ones ie the dialer and contacts, dates... May 04 21:39:58 Yeah but a more complex mechanism would be needed - if you have 200 apps installed but mainly use the non-standard ones then there's no point caching those you're not using May 04 21:40:08 Hm, it's not so much desktop computers as it is wired phones that people are used to, I guess. May 04 21:44:08 SuN: true, if by that you mean, mobiles out on the market at the moment. i had an Nokia N80 a while back and i thought that was slow May 04 21:44:31 my n-gage is slow May 04 21:44:33 not that slow, though May 04 21:44:46 That seems to be a Symbian/Series 60 quirk. May 04 21:44:48 abraxa_: yeah, i guess tagging the programs you use most often, like a rating and cashing the top 5 May 04 21:45:15 SuN: i got rid of it after a while, the N95 is meant to be just as bad May 04 21:45:33 * LarstiQ his siemens 35 happily waits 4 seconds between closing an sms and showing the inbox May 04 21:45:39 mjr: seems to be a common theme with nokia, used to be really good/reliable not slow and crash often May 04 21:46:02 so some slowness I can deal with May 04 21:46:05 mjr: i mean now slow and crash often May 04 21:46:19 A friend of mine has an N73, same thing. May 04 21:46:31 Has to wait about 10 seconds before he can take a pic with his flashy 3MP cam. May 04 21:46:48 LarstiQ: yes some, to a point. and certainly during the dev stages of the neo May 04 21:46:56 If however, you can show someone your neo, and it goes instantly between apps... May 04 21:46:57 The UI has fancy effects though, but I'll take my E6's simple but quick and effective file manager over it any time. May 04 21:47:08 That would be a major plus. May 04 21:47:16 webjames: I wouldn't say no to speed, that's true :) May 04 21:47:18 There's a lot that's wrong with mobile phone software nowadays. May 04 21:47:21 SuN: exactly, no point really, by the time the camera loads up the persons got bored, or the sun has set May 04 21:47:41 Yeah, or the celebrity, random cute chick or fancy car has disappeared. May 04 21:48:22 SuN: sounds like someones talking for experience May 04 21:48:29 :D May 04 21:48:51 It's more or less what's going around in today's youths. May 04 21:49:09 LarstiQ: i think speed is good, but not at the cost of efficiency May 04 21:49:28 webjames: efficiency of what? May 04 21:49:51 webjames: agreed. But at the cost of fancy visual effects... I'd gladly trade that for speed. May 04 21:50:11 LarstiQ: code, i think the N80 demonstrates bad coding, hence the bad speed May 04 21:50:37 efficient coding and fast hardware = a great phone! May 04 21:50:45 * LarstiQ blinks May 04 21:51:13 I don't see how speed comes at a cost of bad coding here, unless you mean on the maintainability side? May 04 21:51:15 webjames sit's here with beryl flashy cube thing on his desktop May 04 21:51:35 :) May 04 21:51:36 ...actually.. slow SMS are probably not matter of bad coding. Remember, that sms-es are stored on sim card -> slow. May 04 21:51:55 SMS'es don't necessarily have to be stored there, I know mine aren't. May 04 21:52:16 pavelm: no non modern phones, i think 10 are then they get stored on my phones memory May 04 21:52:20 most modern handsets don't store them there at all May 04 21:52:33 (the sim card that is) May 04 21:52:57 nor contact for that matter May 04 21:52:59 web: yes, but that's why the siemens was slow. It had to retrieve the first ten from sim every time. May 04 21:53:00 Same for contacts, since the sim card doesn't even allow for more than 10(maybe more, maybe less) characters and no separation between last and first names, let alone other data. May 04 21:53:21 aye, the metadata the sim card can store is very limited May 04 21:53:33 it's a huge hassle when changing phones ;) May 04 21:53:49 pavelm: i think that's something to remember with the neo, all sms and contacts in memory not on sim May 04 21:54:04 web: That's _not_ right solution. May 04 21:54:10 cached in memory May 04 21:54:12 Yeah, a unified bluetooth sync or some other sort of migration system is something that would be useful. May 04 21:54:37 sun: That's called syncml. May 04 21:54:42 pavelm: sorry not memory in the flash May 04 21:54:48 I mean from mobile to mobile. May 04 21:55:04 speed: Yes, cached in memory is good. But please keep copy at simcard so that they are not lost when I change phone. May 04 21:55:05 sun: i use MultiSync May 04 21:55:10 yeah. May 04 21:56:02 thing is i almost have a profile for each person, birthday, email, office address, etc... May 04 21:56:11 so the sim just isn't good enough May 04 21:56:19 a copy on sim is good to keep, but also keep in mind that the copy on sim just won't be full fidelity if you want any sort of metadata there May 04 21:56:22 yeah May 04 21:56:42 I thought I'd migrate my contacts using the sim from my old Moto E680... turned out I had to edit every single one of them. May 04 21:56:58 Hm, apparently my E6 should support SyncML May 04 21:58:01 I wonder why I haven't heard of anyone using it. May 04 21:58:17 * SuN waves May 04 21:58:25 SuN: it's probably buggy :) May 04 21:58:32 SuN: lots of SyncML implementations are May 04 21:58:34 Good chance :) May 04 21:59:01 sorry dropped out there for a moment May 04 22:02:40 the stylus looks pretty fully featured! May 04 22:04:25 hi May 04 22:04:36 Hello r3inforce May 04 22:19:13 http://fun.drno.de/pics/ronald_geht_fremd.jpg May 04 23:16:49 counter May 04 23:16:49 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, 3 days 00:43:09 (10.030 +-3.5 days) (1091;194) May 04 23:22:12 elrond... I wonder why you still try to ask the counter? May 04 23:22:50 pavelm: hope :-D May 04 23:23:22 I know hope is nice... but it seems pretty clear that phase 1 will be in september... May 04 23:23:25 ...2018... May 04 23:23:56 pavelm: its not that far away :-) May 04 23:24:09 We won't need phones by then anymore - we'll use in-brain communicators ;) May 04 23:24:30 (Ok, that was unnice. But are there any better estimates?) May 04 23:24:49 pavelm: yes but Im under NDA May 04 23:25:03 abraxa: No way I'd allow communicator near my brain. May 04 23:25:06 Factory runs are on the 10th and 20th - add a couple of days and you get an estimate May 04 23:25:34 I have birthday at may 18... phase1 openmoko would be nice gift ;-) May 04 23:26:41 Is there some GPS app in the works? I have bluetooth gps here... having moving map would be nice. May 04 23:27:11 I guess gpsdrive could be ported... May 04 23:30:21 pavelm - I asked it to distract me from something. It was completely useless, I know. May 04 23:32:14 pavelm - A friend and I play around with python-gtk stuff. ;o) May 04 23:34:57 elrond: Interesting :-). Python-gtk would mean usable gui scripting. May 04 23:35:07 ...anyway, I have to go to sleep, now. May 04 23:35:12 ...bye bye. May 04 23:36:48 I do have to admit, though, that using python-gtk won't help performance and if performance is already in dire need of optimization... ;x May 04 23:37:00 pavelm - Good night. :) May 04 23:37:32 pavelm - But python-gtk is slow for map drawing, I presume, we're currently hacking tricks out of python for simple drawing stuff. ;o) May 04 23:37:59 abraxa_ - Yep. It's more of a fun project. "hack away" ;o) May 04 23:39:00 A propos hack away... were you around when the Braunschweig Connection (TM) announced that it would be present at the CCCC later this year? May 04 23:39:31 It makes me want to go and in contrast to linuxtag it's gonna be a lot cheaper, too =) May 04 23:40:35 Where's the CCCC? May 04 23:40:54 And no, I weren't around, when they announced it. May 04 23:41:00 http://events.ccc.de/camp/2007/ May 04 23:41:36 http://events.ccc.de/camp/2007/GSM_Village - look, OpenMoko! =D May 04 23:44:12 Ahh. May 04 23:44:50 I consider linuxtag also because some women I wanted to meet since a year gives a talk. ;) May 04 23:45:03 *laughs* May 04 23:45:34 Now that's interesting since there sure aren't too many women able to give talks at such techy meetings May 04 23:45:57 able? May 04 23:46:39 able in the sense of "knowledgable enough" May 04 23:46:45 abraxa_ - Don't let them hear that. ;) May 04 23:46:48 you think of women as in some way less capable of doing such techy things and knowing about these things? May 04 23:47:13 * Elrond gets the feeling, that guaqua is a women ;o) May 04 23:47:18 No but the male-female ratio isn't anywhere near 50:50 May 04 23:47:31 Elrond: that's actually not true, i'm quite male May 04 23:47:46 abraxa_: your choice of words wasn't very good May 04 23:48:10 talking about ability when you are actually talking about ratios and quantities May 04 23:48:44 guaqua - :-) May 04 23:49:05 Well, if they're not familiar with such techy subjects then they're not able to talk about them - that's nothing gender-specific May 04 23:49:29 abraxa_: and in the same sentence you generalized the word ability with a gender May 04 23:49:36 And since there are less women familiar with linux internals than there are guys... May 04 23:49:36 abraxa_ - It's a matter of choosing words. May 04 23:50:00 it's a fine line - whether you sound like a chauvinist or not May 04 23:50:14 I guess I did word it badly then May 04 23:50:20 methinks guaqua spends too much time trying to be politically correct May 04 23:50:26 guaqua: You should consider that not everyone is a native english speaker... May 04 23:50:31 abraxa_: i do know you didn't mean it that way May 04 23:50:51 Good, then I'm glad May 04 23:51:18 Kensan: does it make a difference? if someone finds it degrading it doesn't mean whether the other party actually meant it May 04 23:51:25 And Elrond successfully distracted me from asking who he wanted to see. Smart move ;) May 04 23:51:50 guaqua: yeah it does. If people don't mean offense then none should be taken. May 04 23:52:37 pjz: how can you tell if no-one acknowledges it? May 04 23:52:49 guaqua: how can you tell what? May 04 23:53:15 pjz: how can you tell they didn't mean bad by it? May 04 23:53:16 *lol* May 04 23:53:20 abraxa_ - I wont tell you. May 04 23:53:59 Elrond: I figured as much - not like it makes any difference since I won't be there :) May 04 23:54:16 guaqua: you should presume they didn't mean bad and maybe privately afterward ask them if they meant it in the way you consider bad May 04 23:54:36 pavelm atm i'm pretty confident it does not take that long till p1. http://lists.openmoko.org/pipermail/announce/2007-April/000012.html has some recent dates which i am quite confident in (still) May 04 23:54:42 pjz: Unfortunately it's the other way around usually ;x May 04 23:55:02 guaqua: I said what you *should* do, not what people usually *do* do May 04 23:55:08 pjz: imagine the situation where there's a woman hacker in a new irc channel / similar communication media May 04 23:55:21 sure May 04 23:55:29 pjz: is she going to go ask privately about someone's hasty remark? May 04 23:55:40 guaqua: if it's important to her, she might May 04 23:55:48 guaqua: or... she might just not care May 04 23:56:38 well that most likely isn't the case May 04 23:56:55 guaqua: well, look at the other side; I could choose to deem anything you say offensive, for any (or no) reason whatsoever May 04 23:56:56 she probably cares, but probably also looks for another group of people to collaborate with May 04 23:57:10 guaqua: does that mean you're not going to talk? May 04 23:57:16 err? May 04 23:57:28 well i can deem you are an chauvinist ass, too May 04 23:57:41 guaqua - Or start's ignoring that person. :) May 04 23:57:44 guaqua: sure you can! you can be wrong, too :) May 04 23:58:02 but i'm not doing that - instead i'm trying to convince you of the fact that being politically corret is important May 04 23:58:39 being politically correct does more harm than good, because it discourages straight communication May 04 23:59:08 it doesn't May 04 23:59:40 sure it does May 04 23:59:51 define the ways it does that May 04 23:59:56 it slows down communication, at the very least May 05 00:00:12 i'm not swallowing "sure" May 05 00:00:12 slows down, how? May 05 00:00:13 guaqua - But you're quite right... thinking about things in the past: I took part in a fork, just because the main developers and I (we) didn't talk the same language. May 05 00:00:44 pjz - It slows it down in a good way. May 05 00:00:57 pjz - People don't get angry too fast. May 05 00:01:11 A little slower in communication usually is good. May 05 00:01:16 being straight is hoped for, not being an ass May 05 00:01:57 guaqua - It takes a little to phrase a sentence in a political correct way. :) May 05 00:01:58 right, being straight is hoped for, and if I have to bend my language in knots to avoid offending some sensitive soul(s), then my communication can't be straight May 05 00:02:15 BUT: While phrasing it, one thinks again about the matter. Which is always good. May 05 00:02:33 i just wouldn't like to see this community lose any enthusiastic female hackers because of insensitivity and bad choosing of words - generally easy things to avoid May 05 00:02:40 the problem is who gets to determine what PC language is 'acceptable' May 05 00:03:17 i'm not talking about PC language in general May 05 00:04:03 how would you find a sentence: "not many latinos are able to do such techy things" May 05 00:04:18 is that politically correct? May 05 00:04:32 sure May 05 00:04:42 it's wrong, but it's not offensive May 05 00:04:53 wrong in what sense? May 05 00:04:58 there's plenty of mexican & south american hackers May 05 00:05:21 guaqua: Actually that could be just stating a fact if the majority of latios lacks the knowledge/education to do so May 05 00:05:22 there's also an implication of latinos being less able to do those things in general May 05 00:05:40 guaqua: I disagree May 05 00:05:44 hmm May 05 00:05:49 maybe that was a bad example May 05 00:06:02 the original sentence had a different sound, in my opinion May 05 00:06:25 guaqua: I don't think that implication is there May 05 00:06:30 To me it's a delicate subject either way :) May 05 00:06:44 anyway May 05 00:06:58 just think before you say anything May 05 00:07:05 and the solution to it being delicate is to not *treat* it as delicate by not reading more into it (via implication, for instance) than it actually says May 05 00:07:36 pjz: if you have a thick skin it doesn't mean the others have May 05 00:08:03 guaqua: so we should feed the men babyfood because the children can't eat steak? May 05 00:08:06 pjz: it doesn't give you the right to treat them the way you yourself can be treated May 05 00:08:32 pjz - If more children are around: Yes. May 05 00:09:13 Elrond: maybe so. But then, by your own lights, you're being discriminatory against the men! May 05 00:09:48 off to sleep May 05 00:10:12 nighty night, keep the big wheel rolling May 05 00:10:19 guaqua: g'night! May 05 00:10:21 Night night May 05 00:10:33 pjz - Then get babyfood and steak out. ;-) May 05 00:10:46 guaqua - Night. :) May 05 00:12:18 Elrond: heh. My point is just that I think people shouldn't be castigated for the implications someone chooses to read into their statement because anyone can choose to read anything into any statement. May 05 00:13:25 pjz - You are _also_ right. May 05 00:14:32 When offended people complain, people should start to think over their peach, though. May 05 00:16:34 sure May 05 00:16:47 but to get back to the impetus of this... did anyone complain? May 05 00:17:10 guaqua complained a little. ;-) May 05 00:17:39 pjz: Hard to tell - maybe if a girl/woman would've been around she would've read and feeling offended without saying anything May 05 00:17:45 Though that's just pure speculation again May 05 00:18:36 pissing of (good) people is too easy. May 05 00:18:45 I saw it happen too many times. May 05 00:20:35 hrm May 05 00:20:38 anyway May 05 00:20:52 anyone want to help me debug a bug in the contacts app? May 05 00:20:54 And really: One of the really good things of openmoko is the good atmosphere in here. May 05 00:21:24 Elrond: Yeah, we get free drugs here :) May 05 00:21:38 pjz: What bug? May 05 00:21:49 Elrond: right, and one thing I like about it is that people don't get stomped on for saying something that sounds wrong even when they don't mean it that way :) May 05 00:22:07 pjz - Yeah. :) May 05 00:22:16 abraxa_: if you're editing a contact and add another field of any type, you can't then remove that field May 05 00:23:08 abraxa_: it's in contacts-contact-pane.c May 05 00:23:35 abraxa_: field_button_remove_cb May 05 00:23:39 So don't get this wrong now: I need sleep. :) May 05 00:23:49 Elrond: heh, g'night! May 05 00:24:13 abraxa_: the remove variable in that routine is *never* true May 05 00:24:35 abraxa_: it's always showing up as false and just clearing the text May 05 00:24:55 pjz - If you don't find the bug until in 18 hours, I'll go and take a deep loo at the code too. :) May 05 00:25:26 So, good night people! :) May 05 00:26:11 Unfortunately I'm not familiar with that piece =x May 05 00:33:27 my problem is I really don't know how to build it for x86 so I can test it May 05 00:34:57 morning May 05 00:35:06 howdy May 05 00:35:23 rwhitby: how do I use bitbake to build just the openmoko stuff for x86? May 05 00:35:52 make setup-machine-x86 openmoko-devel-image May 05 00:36:01 rwhitby: (right now MokoMakefile is dying trying to build the entire image for x86 due to some kind of bug in libelf May 05 00:36:31 ah, yes. nothing I can do about that stuff in mokomakefile. I get the same problem but don't know how to fix it. May 05 00:36:40 yeah May 05 00:37:00 so how do I bypass mokomakefile and use bitbake directly to build just the openmoko apps? May 05 00:37:21 make package- I think May 05 00:37:39 or "cd build ; . ../setup-env ; bitbake " May 05 00:37:51 (the lone dot is important) May 05 00:38:31 but that will still probably depend on libelf. May 05 00:38:42 (if done with bitbake and OE) May 05 00:38:49 pjz: What bug about libelf are you getting? Maybe it's the one that's in bugzilla... May 05 00:39:15 If you're really asking about how to use qmake or autconf to build the openmoko apps from svn only without bitbake or OE, then I haven't set anything up for that yet. May 05 00:41:01 * rwhitby was on holidays for two weeks, then setting up a US office network for this week, then will be in .eu on business next week. May 05 00:41:10 gotcha May 05 00:41:26 you were just the one person here I knew was an OE hacker to some extent May 05 00:41:30 I know zero about OE May 05 00:41:41 I don't even have my P0 phone at the moment - I gave it to shenki to play with while I knew I was going to be too busy to do anything with it. May 05 00:44:31 hrm, maybe I'll just have to hack together a bog-standard Makefile buildsystem for OM-2007 May 05 00:45:24 A while ago I compiled the libs here as well - only using the shipped makefiles May 05 00:45:35 oh, using the qmake stuff? May 05 00:45:38 So maybe you won't even need that May 05 00:45:51 I think so, it's been a fair bit :) May 05 00:46:35 pjz: it should be easy to use the existing autoconf stuff in the svn. if someone works out how to do it I'll be happy to add it to mokomakefile May 05 00:46:53 Just I didn't know how I could have them put the resulting binaries *not* into /usr... May 05 00:50:15 glib-genmarshal --prefix=moko_gsmd_marshal ${QMAKE_FILE_IN} --header --b May 05 00:50:19 ody >/moko-gsmd-marshal.h May 05 00:50:20 that's... weird May 05 00:50:27 putting a .h in / May 05 00:50:54 undefined variable? May 05 00:51:08 no, that line is from the Makefile May 05 00:51:36 oh, that's just plain wrong then May 05 00:51:44 that's what I thought :) May 05 00:51:53 should it be a >> ? May 05 00:52:04 though... it could be an undefined var in whatever qmake uses to build the makefiles? May 05 00:52:17 or... hrm, there's a Makefile.am in there May 05 00:53:11 yeah, undefined variable 'srcdir' May 05 00:56:29 this effort clearly needs beer May 05 00:56:34 lol May 05 00:56:45 * pjz drinks some Golden Monkey May 05 00:57:48 I just tried building the libs and it works so far May 05 00:58:06 Used supplied autogen.sh, then qmake/make in each lib dir May 05 00:58:58 oh, interesting May 05 00:59:01 trying that May 05 00:59:30 Just.. hm.. the libs depending on glib/gtk can't find it May 05 01:02:47 apt-get install ... May 05 01:03:17 I have both but the include paths don't get send to gcc May 05 01:03:22 sent* May 05 01:03:56 hrm May 05 01:06:59 Any progress on your side? May 05 01:07:07 a little May 05 01:07:17 running autogen.sh multiple times seems to help May 05 01:07:19 somehow May 05 01:08:40 I installed libecal and some other libs it wanted May 05 01:09:01 checking for OMOKO... configure: error: Package requirements (openmoko-libs) were not met: May 05 01:09:04 No package 'openmoko-libs' found May 05 01:09:06 Consider adjusting the PKG_CONFIG_PATH environment variable if you May 05 01:09:09 installed software in a non-standard prefix. May 05 01:09:17 what should PKG_CONFIG_PATH be? May 05 01:09:44 If memory serves me well it's the path where pkgconfig can find the .pc files May 05 01:10:10 hrm. there aren't any of those. May 05 01:11:27 How did you resolve the libgsmd dependency? May 05 01:11:34 I didn't :P May 05 01:11:37 lol May 05 01:12:22 I switched over to trying to build the app May 05 01:13:24 Actually it should be easier to just have MokoMakefile build the entire rootfs for x86 and just chroot into it May 05 01:18:54 ...except for the bit where it errors out due to libelf or something May 05 01:19:45 * pjz reruns mokomakefile to repro the error May 05 01:19:53 * abraxa_ nods May 05 01:26:23 I think this counts as 'too difficult' for your average wannabe contributor May 05 01:26:38 "this"? May 05 01:27:06 the openmoko tree in svn May 05 01:27:19 and the difficulties we're having trying to just get it to build May 05 01:27:56 Indeed, but I'm not fussed - it's to be expected at this stage with so few people dedicated to the project May 05 01:28:02 I guess May 05 01:28:17 I'm 99% your libelf issue is this btw: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=264 May 05 01:28:26 99% sure that your... May 05 01:29:09 yup, that's it May 05 01:30:13 Good :) May 05 01:30:24 Then at least you know how to get around it May 05 01:31:07 yup May 05 01:31:12 cool, thanks for pointing that out May 05 01:32:00 Yupyup, it's what this channel is for ;) May 05 01:33:11 And with that I'm off to bed as well... nighty night May 05 01:34:03 gnite! May 05 01:34:05 sleep well! May 05 01:34:39 You, too, whenever you do :) May 05 01:35:42 still early here in TX :) May 05 02:56:29 openmoko: 03laforge * r1910 10/trunk/src/target/u-boot/patches/uboot-20061030-neo1973.patch: * make LCM SPI timings slower (200uS clock cycle rather 150uS, 90mS Top instead of 50mS) to work around bad timing info in data sheet **** ENDING LOGGING AT Sat May 05 02:59:56 2007