**** BEGIN LOGGING AT Fri Jun 15 02:59:56 2007 Jun 15 04:26:58 * * OM Bug 607 has been created by roh(AT)openmoko.org Jun 15 04:26:59 * * flash.sh stalls Jun 15 04:27:00 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=607 Jun 15 04:48:24 ImportError: No module named bb Jun 15 04:50:22 what has changed in recent days? Jun 15 04:50:22 i can't build my openmoko now Jun 15 05:42:14 heya Jun 15 05:42:29 Mornin Jun 15 06:17:30 morning Jun 15 06:28:30 Readon: your bitbake installation got broken Jun 15 06:39:29 hrmm... any updates on when i can give my $350 Jun 15 06:39:30 lol Jun 15 06:39:55 i keep seen iPhone advertisements and i want to devel my openmoko to be even cooler ;-) Jun 15 06:40:27 get both :P Jun 15 06:40:52 If I can come up with the money, I will probably do something along that line Jun 15 06:41:09 OpenMoko development seems very tempting Jun 15 06:41:10 daMaestro: possibly news any day now Jun 15 06:41:31 meh.. i don't want to spend any money on the iPhone Jun 15 06:41:50 their dev kit consists of "oh .. just use safari and all the neat new web 2.0 technologies!" Jun 15 06:41:50 lol Jun 15 06:41:56 blah. Jun 15 06:42:01 closed. platform. Jun 15 06:42:06 yeah, I am a bit disappointed there Jun 15 06:42:24 Hopefully Apple will at least do a "select" dev program Jun 15 06:42:35 get your app certified sort of thing Jun 15 06:45:35 supposedly QA has been done by now Jun 15 06:46:04 yeah.. i really just want a phase 2 device now Jun 15 06:46:05 lol Jun 15 06:46:24 but i'm willing to buy a phase 1 device to play with if i get some sort of decent discount on the upgrade Jun 15 06:47:05 anyone know what media management software is slated? Jun 15 06:47:25 i assume something light such as listen would run? Jun 15 06:47:41 * daMaestro has not done much research... Jun 15 06:47:55 i'll just get too excited and waste time reading every single spec on the hardware, etc Jun 15 06:49:23 aloril: I updated you wiki-counter-message this morning Jun 15 06:49:27 counter Jun 15 06:49:28 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 55%) P1: 3 days 14:35:14 (3.608±3.6 days) (1472;228) Jun 15 06:50:22 Not much, but at least that Harald is back tomorrow, hopefully with good news. :) Jun 15 06:50:42 CM: thanks Jun 15 06:51:05 From 04:11 this morning :) Jun 15 06:52:26 so its 1 more day more than I though, news today seems unlikely then Jun 15 07:02:11 anyone having problems with flashinglately Jun 15 07:05:42 Sup3rkiddo on qemu or at real hardware? Jun 15 07:06:06 roh, on qemu...flashing the local image Jun 15 07:06:14 thats a bit broken Jun 15 07:06:19 getting a timeout Jun 15 07:06:47 hmmm....think a bug should be filed? Jun 15 07:08:24 use a qemu from before r2198 Jun 15 07:08:48 i already files a bug. you can also compile a new qemu without unbusy-i2c.patch Jun 15 07:08:52 good morning Jun 15 07:09:16 roh, oo....can you tell me how can i revert back to an older qemu..? Jun 15 07:10:06 Superkiddo: just move all newer versions out of build/tmp/deploy/images/ Jun 15 07:11:43 oh ok....but the older guys have the problem as well....built from scratch yesterday Jun 15 07:11:59 Sup3rkiddo if you don't have older builds flying around,take the 2157 from http://buildhost.openmoko.org/tmp/deploy/images/ Jun 15 07:12:30 these should work. 2157 is the one before the change Jun 15 07:12:54 roh, but i have some modification to the build... Jun 15 07:13:15 to u-boot? Jun 15 07:13:26 roh..ok..no..i get it nm Jun 15 07:13:28 *now Jun 15 07:18:37 So, I take it that OpenMoko can be run in qemu then? Jun 15 07:19:39 dkirker, yup Jun 15 07:19:56 I am looking at the Wiki, are there any good instructions? Jun 15 07:20:16 dkirker, wiki is as good it gets :) Jun 15 07:20:21 *as Jun 15 07:20:41 ok, I'll poke around. Jun 15 07:21:15 I am working on a mobile browsing/messaging/synchronization client core, and I am now very interested in porting it to the OpenMoko platform. Jun 15 07:21:44 roh, works like a charm...thamk you Jun 15 07:21:47 *thank Jun 15 07:22:21 hi jserv2 Jun 15 07:23:20 Sup3rkiddo: ok, I am seeing instructions and links now! Thanks! :) Jun 15 07:23:34 HTH :D Jun 15 07:25:57 morning Jun 15 07:26:13 morning Jun 15 07:26:16 counter Jun 15 07:26:17 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 55%) P1: 3 days 14:16:49 (3.595±3.6 days) (1473;228) Jun 15 07:36:41 openmoko: 03mickey * r2264 10/trunk/oe/packages/ (24 files in 11 dirs): oe: add infrastructure tweaking to make core fonts display w/ kdrive Jun 15 07:40:33 counter? Jun 15 07:40:33 "/msg aloril counter?" for actual long message, giving short version here: http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 55%) P1: 3 days 14:09:41 (3.590±3.6 days) (1474;228) Jun 15 07:57:19 CM: added 2 other items from IRC too (still catching up on logs) Jun 15 07:57:41 aloril: Nice work, I like that counter page Jun 15 07:57:47 Good to have when you're busy for a few days Jun 15 07:58:19 CM: thanks, what I'm most worried is that somebody takes rumor/guess too seriously Jun 15 07:59:04 what? like demanding that i should have been able to buy a device .01 days ago! ... aloril Jun 15 07:59:04 lol Jun 15 08:00:11 daMaestro: hehe, more like reading "Guess: some foo feature for bar device" and then assuming that .. Jun 15 08:00:27 ahh Jun 15 08:00:48 yea.. i really want phase 2.. but i am going to get a phase 1 device to see what applications i can port over Jun 15 08:01:02 so what package management is planned? Jun 15 08:01:19 is it something homebrew? is there even one? Jun 15 08:01:31 does anyone have a spare device...send it to me...pleeeeaase Jun 15 08:01:36 "P Jun 15 08:01:55 Anyone can buy for $350 starting perhaps any week now Jun 15 08:02:04 so can I already buy one device? Jun 15 08:02:11 is it stable yet? Jun 15 08:02:28 I had a bad experience with qtopia greenphones... Jun 15 08:02:33 am not sure about the stability part Jun 15 08:03:04 Zenton: you can get OpenMoko now, device hopefully soon Jun 15 08:03:10 faq? Jun 15 08:03:10 See http://wiki.openmoko.org/wiki/FAQ Jun 15 08:03:16 thanks aloril Jun 15 08:03:28 (has links to OpenMoko, etc..) Jun 15 08:04:37 daMaestro: It uses ipkg Jun 15 08:06:17 Zenton: hardware or software problems? Jun 15 08:06:36 mostly hardware Jun 15 08:06:53 batteries gave up working. Jun 15 08:07:40 Zenton: for Neo1973 you can use ubiquitous Nokia batteries too (model is mentioned in Wiki somewhere) Jun 15 08:08:05 so even if Neo1973 has similar problems, you can cheaply get replacements ;-) Jun 15 08:08:18 aloril: I gonna buy a Neo1973 as soon as I can. Jun 15 08:08:29 -"- Jun 15 08:14:01 the problem with BL-5C is that they currently don't charge in the Neo Jun 15 08:14:08 so you need something like DT14 or so Jun 15 08:14:21 (Nokia standalone battery charger) Jun 15 08:14:38 i don't know whether this is a hardware limitation or something that could be fixed in software Jun 15 08:21:35 CM, so.. is there such think as a "repo" or "source" for ipkg's? Jun 15 08:26:17 mickey_fic: thanks, updated http://wiki.openmoko.org/wiki/Neo1973_Battery Jun 15 08:27:19 another option for on the road additional energy: http://www.ladyada.net/make/mintyboost/ Jun 15 08:27:34 as far as I know, what is needed in addition to above: Jun 15 08:28:10 that's fun Jun 15 08:28:21 1) configure Neo1973 to charge using 500mAh instead of default 100mAh (just run suitable software I assume) Jun 15 08:28:46 2) suitable cable (maybe comes with Neo1973 already?) Jun 15 08:29:31 UI question: what's your preferred position of short-lived OSD text messages? Jun 15 08:29:38 center, bottom, top? Jun 15 08:29:43 left aligned, center, right aligned? Jun 15 08:29:49 kick-ass large / small? Jun 15 08:29:57 [as in font] Jun 15 08:30:02 right-bottom-small Jun 15 08:30:33 *right-top Jun 15 08:31:35 CM, something like "yum repo" or "apt source" for ipkg? Jun 15 08:31:43 i just briefly read about it Jun 15 08:35:16 bottom-left-big (just the opposite in every way ;-) Jun 15 08:35:44 mickey_fic: Do you know how long can the device stay on with a charged battery ?? (lcd, ts, cpu etc on, but not phoning) Jun 15 08:36:39 mickey_fic: top, right aligned (e.g. this way I have "Konversation" configured if you do "schurig: BLAH") Jun 15 08:37:02 aloril: eh.. doesn't this schematic only provide 5V 100mA? Jun 15 08:37:06 schurig: I had the p0 running for about 20000 seconds at full backlight brightness, writing to microsd every 10 seconds Jun 15 08:37:06 daMaestro: cat /etc/ipkg.conf :-) Jun 15 08:37:10 schurig: gta01 3 to 4 hours or so Jun 15 08:37:17 mickey_fic: I know you are busy, but once you (==FIC in general) are not that busy: maybe something like mintyboost but one that can also act as USB hub (giving power) to 1-4 devices (I remember Sean mentioning plans about something like this some time ago in ML) Jun 15 08:37:19 schurig: we never measured that Jun 15 08:37:48 pH5: that's 13 hours. Not bad. Jun 15 08:38:01 pH5: really??? Jun 15 08:38:02 pH5: with LCD auto-dimmed? (which is OK, common use case) Jun 15 08:38:22 roh: no, but I think Neo1973 by default won't draw more than 100mAh -> thus need to have software where this is overridden (mintyboost needs some changes to so that Neo1973 knows it can draw 500mAh like newer Neo1973 charges has I think) Jun 15 08:38:26 pH5: ah, sorry, you wrote "full backlight brightness", confused your entry with mickey_fic .... Jun 15 08:38:27 schurig: that's 5.5 hours, no backlight dimming, but the cpu was idle most of the time Jun 15 08:39:02 aloril: i know there's demand. once gta01 is out, gta02 is further, we start thinking about peripherals Jun 15 08:39:03 aloril: sadly to say, it draws more. it even discharges the battery when only fed with 100 Jun 15 08:39:26 with everything on and on full power Jun 15 08:39:28 schurig, i don't have a device or a devel env (virtual) setup Jun 15 08:39:32 schurig, :-/ Jun 15 08:40:05 mickey_fic: http://www.flickr.com/photos/ph5/422949345/ Jun 15 08:40:40 mickey_fic: hehe, some of your problems sound like our (early) problems with Ramses or RT4000 :-) Jun 15 08:41:42 mickey_fic: because we had the same problem: device turned on, put in the docking cradle or vehicle dock, and yet the battery was slowly discharged. Our hardware guy fixed this a week later :-) Jun 15 08:42:21 mickey_fic: but (I've seen the schematics), we have a VERY complicated analog part for power. Lots of individual component, no charging controller IC or so. Jun 15 08:43:37 schurig: when the usb does not speak usb but only provides power a usb device is not allowed to draw more power than 100mA . when your device draws more you cannot do anything besides violating the usb standart. thats why we use a id pin now to detect dumb loaders able to supply 500mA Jun 15 08:44:08 roh: maybe that's the reason why now in the Linux kernel is a dummy USB charger driver Jun 15 08:44:28 pH5: nice! Jun 15 08:45:07 roh: see CONFIG_USB_BERRY_CHARGE in 2.6.21 Jun 15 08:46:01 schurig oh.. didn't see that. the moko is been detected as usb-acm in uboot and as cdc-ethernet when the kernel is up. afaik linux supplies every device, if a driver is loaded or not woth 500mA if it asks nicely. windows doesnt. it even poweres off ports actively to enforce that Jun 15 08:47:07 pH5, you guys have all the fun...mwaaaa Jun 15 08:47:17 roh: hmm, this (black)berry usb charge driver is somewhat different. It sends an via USB a command to the device. "The device will then reset itself with the new power usage and should start charging" Jun 15 08:48:33 http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=drivers/usb/misc/berry_charge.c Jun 15 08:50:31 roh: you mean that even if I configure Neo1973 to draw 500mAh, mintyboost won't give it (ie.. it only will give max 100mAh)? Jun 15 08:51:18 roh: ie.. even if same kind of 'id pin' was added to mintyboost? Jun 15 08:52:34 aloril at least thats what the schematics and text say as far as i have read Jun 15 08:53:09 aloril: then the moko will draw more power and propably overload the booster which in the best case just goes off Jun 15 08:53:49 good morning Jun 15 08:54:55 hmm.. saw mention of 200mAh there too, would that be enough to use Neo1973? Jun 15 08:55:30 * CM adds another vote to the small upper right corner OSD popups Jun 15 08:55:43 ah, faq says up to 250mA+: http://www.ladyada.net/make/mintyboost/faq.html Jun 15 08:56:30 aloril: it could keep it on, but not really charge anything much, if the voltage keeps high enough at all Jun 15 08:56:57 daMaestro: Sorry, don't know. Haven't even used ipkg myself yet Jun 15 08:59:38 roh: OK, anyway I don't see any other option for "additional usage" when not near power outlet (except to get extra batteries, but I think this needs powering Neo1973 down and I dislike rebooting) Jun 15 08:59:49 well.. charge it from laptop ;-) Jun 15 09:02:30 openmoko: 03mickey * r2265 10/trunk/src/target/OM-2007/openmoko-libs/ (ChangeLog libmokoui/moko-banner.c): libmokoui/moko-banner.c: tweak look and feel Jun 15 09:03:04 aloril: hehe.. guess what i use as charger for my motorola :) Jun 15 09:03:41 CM, i'm going to look at porting rpms to ipkg and doing mass builds of fedora based packages Jun 15 09:04:05 ok, for now I'll leave the OSD centered, but what I've heard from you means it needs to be configurable anyway ;) Jun 15 09:04:08 too many tastes Jun 15 09:04:26 mickey_fic: yes, please :) Jun 15 09:13:35 LaF0rge: is anybody currently working on SMS functionality in gsmd? Jun 15 09:15:15 is there any easy way to check if my notebook outputs 100 or 500ma on usb ports? Jun 15 09:17:14 mickey_fic, so.. where can i send my money? Jun 15 09:17:17 :-P Jun 15 09:20:55 mickey_fic: maybe it doesn't need to be configurable Jun 15 09:21:12 mickey_fic: my preference came from a desktop system, the Neo1973 LCD is much smaller Jun 15 09:26:41 pH5: it's currently on hold -- however there shouldn't be much to add according to LaF0rge. would you like to have a look? Jun 15 09:26:56 aloril: you can still use something like this http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1831 Jun 15 09:27:04 just a bit more power Jun 15 09:30:33 mickey_fic: yes, I've seen the stubs in sms_cb.c and the four list/read/send/write commands in libgsmd-tool. But I'm not sure how it is supposed to be wired together. Jun 15 09:38:05 hmm, I can't get the dialer to display the moko stock icons. I have installed the openmoko-icon-theme - where do I make sure that it is selected by gtk? Jun 15 09:46:08 roh: yeah, looks enough, I guess would require new PCB and possibly other changes though Jun 15 09:54:47 pH5 - BTW: my python thingy got hh-packet-dumping (like your decode.sh) and LE uint32 dumping yesterday ;o) I hope to play with one channel this evening. ;) Jun 15 09:58:35 Elrond: I'm looking forward to it :) Jun 15 10:00:58 * * OM Bug 608 has been created by stefan(AT)openmoko.org Jun 15 10:00:59 * * Kdrive does not supprt option -pf and failed to start Jun 15 10:01:01 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=608 Jun 15 10:01:49 * LuitvD is off for some CoD Jun 15 10:18:57 (script) wiki RecentChanges: [[Wishlist:Text_Input]] [[UI_Improvements]] [[Neo1973_Battery]] Jun 15 10:21:58 * * OM Bug 609 has been created by stefan(AT)openmoko.org Jun 15 10:21:59 * * Kdrive does not supprt option -pf and failed to start Jun 15 10:22:01 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=609 Jun 15 11:04:52 morning all Jun 15 11:08:56 have anybody normal icons in "Today" app (running in qemu)? Jun 15 11:24:33 Hello, everyone. Jun 15 11:25:10 Is there some way to make openmoko ring on incoming call? I see a dialog, but no sound :-( Jun 15 11:28:11 pavelm: there should be special alsa setting Jun 15 11:28:17 somewhere.... Jun 15 11:31:37 magon: aplay produces sound... so speaker seems set up correctly. Jun 15 11:32:07 magon: and it does not vibrate, either. Jun 15 11:35:16 pavelm, maybe this will help http://wiki.openmoko.org/wiki/Manually_using_GSM#Dialling_an_outgoing_call Jun 15 11:40:13 anykey: I'd need "playing sound on incoming call" :-) Jun 15 11:42:03 hey have you lookedat it? Jun 15 11:42:42 s/lookedat/looked at Jun 15 11:42:52 Hi pavelm! :) Jun 15 11:43:05 any: yep. Can't parse libgsm-tool because it buffers its output :-( Jun 15 11:43:10 hi, elrond! Jun 15 11:43:47 pavelm, great :-( Jun 15 11:44:11 elrond: any news? Has your compiling machine finally died because of too much work? ;-) Jun 15 11:44:16 pavelm - BTW pygtk: This means compiling X11R7.3... I stopped that. I suggest you file a wishlist bug on bugzilla.openmoko.org and ask for pygtk. Usually that works. Jun 15 11:44:41 elrond: ok, will do. Jun 15 11:44:42 pavelm - News: My hammerhead python thingy slowly progresses. ;) Jun 15 11:45:11 elrond: wow.., do you get some interesting communication? Jun 15 11:47:36 pavelm - I start to, yes. I just finished the packet extractor-and-dumper yesterday. When I'm getting home today, I'll try to send the basic setup sequence and play with one channel. ;) Jun 15 11:50:16 pavelm - A quick-simple build-for-host script and intro on the wiki is also on my list for today. And finally de-b0rking openmoko-messages ;o) Jun 15 11:50:20 elrond: good luck. Jun 15 11:50:51 I don't know, what I will really do. ;) Jun 15 11:51:05 * pavelm wonders if it would be fun to place packet-modifier between hammerhead and its driver. Slightly tweaking messages and see what happens :-) Jun 15 11:51:09 Anyway. I am supposed to do "real work" now. ;o) Jun 15 11:51:39 so what's this openmoko phone thing ? phones for geeks who want to compile their own software on it ? lol Jun 15 11:52:48 voidwalker: Taiwaneese way of freeing themselves from Nokia control? :-) Jun 15 11:52:55 voidwalker - Currently: Yes. In half a year, it'll hopefully a normal phine which just happens to run opensource software. And which normal people can update just. (and install new software) Jun 15 11:53:10 oh, but don't we already have those ? Jun 15 11:53:16 symbian is not that ? Jun 15 11:53:35 Can you _fully_ control the hardware on a symbian phone? Jun 15 11:53:51 void: Symbian is still nokia's slave :-) Jun 15 11:53:51 i dunno, i never had one Jun 15 11:53:55 oh ok Jun 15 11:54:08 void: Plus symbian does not work. Jun 15 11:54:23 I have no idea on symbian. Jun 15 11:54:34 well it would be cool if this phone had wifi, gps, large memory+memory card, so it can play mp3s, videos etc Jun 15 11:55:00 voidwalker - The mass market version will have exactly all that. Jun 15 11:55:16 at least from current planning. Jun 15 11:55:27 so you don't have to carry a pda, a phone, and an mp3 player Jun 15 11:55:38 I have to go, sorry. Jun 15 11:55:40 bye bye. Jun 15 11:56:34 Yep. And I need to do "real work". Jun 15 11:56:37 Laters. Jun 15 11:56:41 hng, I wish gsmd had the multiline reply support already. it still throws away most of the CMGL, CMGR, etc. answers Jun 15 11:57:32 ;o open source gsm software ? Jun 15 11:57:34 awesome Jun 15 11:57:50 hey, won't this open new ways to hackers who want to cheat the phone company ? Jun 15 11:57:59 the GSM stack is not open Jun 15 11:58:20 well then bye bye free phone concept ;p Jun 15 11:58:23 voidwalker: the gsm stack itself is encapsulated in the gsm chip, we can only talk AT to it. Jun 15 11:58:24 it's on a separate chip, the OpenMoko system accesses it through a serial line Jun 15 11:58:28 oh Jun 15 11:58:48 voidwalker: depends if you want to be able to use it with the GSM network or not :P Jun 15 11:58:54 i don't really know that much about gsm, is the technology closed ? Jun 15 11:58:59 it's suboptimal wrt. freedom, but the best one can do at the moment Jun 15 11:59:47 i mean, what's stopping from making it fully open on this phone ? Jun 15 12:01:14 Playback. Maybe recording Jun 15 12:01:17 only maybe ? Jun 15 12:01:22 First that it's an off-the-shelf GSM chip (or close to it) they got. Jun 15 12:01:43 Second that there are no free GSM stacks and writing one would be a major undertaking. Jun 15 12:01:53 Third that yes, laws are in the way. Jun 15 12:02:32 oi Jun 15 12:02:55 is there still any chance the phase1 phones will be released before mass marketing? :( Jun 15 12:03:03 hell yes Jun 15 12:03:18 i mean like - mass marketing is said to start in around 5 months Jun 15 12:03:36 "is said" Jun 15 12:03:48 Is there any news concerning FCC certification of the Neo? Jun 15 12:04:12 i mean like - whats the reason p1 is delayed that much anyway? Jun 15 12:04:19 does anybody know something? Jun 15 12:04:28 (except sean who doesnt say anything ... at least it seems :P) Jun 15 12:04:47 DukeOfURL: no leaks from the FCC at least Jun 15 12:05:00 they're usually the first ones to blush Jun 15 12:08:38 only 2 buttons ? Jun 15 12:08:54 only 2 buttons Jun 15 12:08:58 and a biiig faaat touchscreen Jun 15 12:09:16 2.8" is not that big Jun 15 12:09:33 get smaller fingers Jun 15 12:09:46 that's a fair cop Jun 15 12:12:31 uh oh Jun 15 12:12:34 it says AGPS Jun 15 12:12:38 not GPS Jun 15 12:12:52 will that work just as well ? Jun 15 12:12:59 will the phone have a gps antenna ? Jun 15 12:13:40 it has a gps antenna Jun 15 12:13:50 agps means gps with prefetched sattelite-position information Jun 15 12:13:53 see wikipedia agps Jun 15 12:14:25 microsd, just 2gb, :| Jun 15 12:14:49 2gb is just not enough Jun 15 12:15:15 depends on what you want Jun 15 12:15:20 you could also use a usb hdd with 500gb Jun 15 12:15:38 Nebukadneza: you don't need A part of GPS ;-) Jun 15 12:15:49 well i'm thinking how well it could replace a PDA Jun 15 12:15:51 "8 mb ought to be enough for anybody" a wise man once said * Jun 15 12:16:04 i have just bought a PDA 2 weeks ago Jun 15 12:16:11 Nebukadneza: also I think further Phase 1 is delayed, further will also mass market be delayed Jun 15 12:16:22 and i'm thinking navigation maps here Jun 15 12:16:45 those take about 1 gigabyte with the software i use Jun 15 12:16:53 about delays: many various reasons (hardware bugs, parts availability) Jun 15 12:16:59 and you would only have 1gb left for mp3s Jun 15 12:17:43 voidwalker: it's not really billed as a music phone though Jun 15 12:17:49 voidwalker: 4gb microsd (are not available yet but) exist and 8gb microsd are announced for 2008 anyway. Jun 15 12:17:55 voidwalker: 1gb of the quality it could pump out would easily outlast the battery imo Jun 15 12:18:17 Nebukadneza: at the moment I think biggest USB powered hard disk is 160GB (though I guess you could always bring big battery and voltage converter to use 500GB one ;-) Jun 15 12:18:50 oh yeah i wanted to ask about that Jun 15 12:18:57 usb powered hard drives Jun 15 12:18:59 160gb you say ? Jun 15 12:19:04 aloril: i doubt that the neo will give enough power to spin a usb powered one ;P Jun 15 12:19:20 neo has unpowered usb Jun 15 12:19:25 so it needs a powered hub Jun 15 12:19:30 thats what i thought Jun 15 12:19:42 Nebukadneza: right, you need external battery powered one (still probably easier than using 500GB ones) Jun 15 12:19:47 basically you'd need to have a battery powered hub to get usb hd working Jun 15 12:20:09 well, powered hub vs. powered hdd is no big difference Jun 15 12:20:38 voidwalker: http://www.lacie.com/fi/products/product.htm?pid=10868 Jun 15 12:22:01 99 EUR in one shop Jun 15 12:23:34 aloril: you could buy a 500GB disk for that amount of money Jun 15 12:24:13 LuitvD: yeah, I wonder what kind of voltage they actually use (would need battery for usb and hard disk separately) Jun 15 12:24:36 LuitvD: also those 500GB hard disks are a quite amount bigger than above product ;-) Jun 15 12:24:42 true Jun 15 12:25:01 you know, there are USB powered mobile harddisks Jun 15 12:25:23 LuitvD: see above link (above 99 EUR price) Jun 15 12:25:26 those have two plugs, of which you plug each one into a different powered hub Jun 15 12:25:35 is that one usb powered? :| Jun 15 12:25:45 yes, one USB port too Jun 15 12:25:52 strange Jun 15 12:26:15 USB powered, 160GB, 135 x 75 x 15 mm Jun 15 12:26:21 meh, I'd rather have a 500GB in a server, and a network connection Jun 15 12:26:29 codafs! Jun 15 12:29:36 what? Jun 15 12:29:38 LuitvD: even with GTA02 and WiFi: you don't have WiFi everywhere ;-) (and even if has EDGE, its too slow) Jun 15 12:30:07 but maybe some future model has enough fast network connection in many places Jun 15 12:30:15 aloril: yeah, EDGE isn't an option here even... so it's just WiFi Jun 15 12:30:45 well i would buy it, but i fear it will cost well over $400 here :| Jun 15 12:30:46 if it's HSDPA 1,8Mbit it's good enaugh for me... Jun 15 12:31:56 HDSPA coverage in NL is getting better Jun 15 12:32:36 though about a year from now, WiMAX might be the nicest option... they're already doing some tests here Jun 15 12:33:14 LuitvD: not only use network storage, but also move computing intensive stuff to server (like kernel compile) Jun 15 12:33:25 or chess/go program or .. Jun 15 12:33:32 yeah Jun 15 12:33:45 mobile phones should become thin clients... Jun 15 12:34:22 * aloril intents to use Neo1973 like thin client, even with GPRS (in addition to using it as small PC too) Jun 15 12:34:36 ie.. some programs locally, other programs remotely Jun 15 12:34:36 buh, thin client over gprs... Jun 15 12:34:48 very thick thin client Jun 15 12:35:11 FreeNX should be usable over modem connections Jun 15 12:35:14 should only sent really really computing intensive operations... Jun 15 12:35:26 s/sent/send/ Jun 15 12:35:27 LuitvD meant: should only send really really computing intensive operations... Jun 15 12:40:19 talking about thin client over gprs... does anyone know about typical latencies over gprs? Jun 15 12:40:44 ping isn't that good, AFAIK Jun 15 12:40:52 0.25-3s Jun 15 12:40:58 that low? Jun 15 12:41:02 :P Jun 15 12:41:05 * aloril has seen 250ms ping Jun 15 12:41:15 I oftenly jump over 3 seconds here... Jun 15 12:41:27 and 250 isn't possible here... Jun 15 12:41:47 humm Jun 15 12:41:52 really, it's quite low where I'm from Jun 15 12:42:04 I usually get 200 ms with hsdpa latency Jun 15 12:43:07 it was faster than GPRS though, EDGE maybe Jun 15 12:44:37 Zenton, typical latencies vary according to local loads rather a lot really... Jun 15 12:45:00 yeah, and operators aren't always very nice about maintaining good latency Jun 15 12:45:17 Zenton: during tests on the Orange network in London last year I was getting 7s round trips Jun 15 12:45:22 I have tried several programs and latencies with hsdpa, now I have to test them with gprs... Jun 15 12:45:24 but let's see Jun 15 12:46:11 * LuitvD wants a HSDPA phone Jun 15 12:46:20 preferably the Samsung F700 Jun 15 12:47:01 currently 600-800 pings at my location, have seen a bit better Jun 15 12:47:03 plain gprs Jun 15 12:47:24 mostly in the 600's, occasional rises Jun 15 12:47:51 with nx and 200ms I can play games like maelstrom. Jun 15 12:48:06 nx maybe, plain X will be death with these latencies ;) Jun 15 12:48:11 but 200ms are not suitable for vnc. Jun 15 12:48:17 HSDPA, 5MP cam, WiFi, wide (2:1 ratio) touchscreen Jun 15 12:48:22 ideal phone for me Jun 15 12:48:25 :) Jun 15 12:49:11 (not to forget the qwerty keyboard) Jun 15 12:49:47 LuitvD: for hsdpa I use a huawei modem. Jun 15 12:50:10 I want it in a phone :P Jun 15 12:50:53 just attach, it to the Neo via USB (plus an external power feed) Jun 15 12:51:02 mjr: X crashes if connections hangs, not usable. Jun 15 12:51:18 Zenton, there's that _too_ Jun 15 12:51:54 Hmm. http://www.ebuyer.com/UK/product/126153/rb/0 Jun 15 13:05:15 Zenton_: doesn't FreeNX support something that allows to continue after reconnecting? Jun 15 13:06:05 yeah, it has like a proxy Jun 15 13:06:56 basically it seems to be a bit like an Xvnc solution, except maybe with a more efficient protocol Jun 15 13:06:59 correct me if I'm wrong Jun 15 13:08:00 it's the X protocol with bits missing, basically Jun 15 13:09:24 the ssh bits of it are a bit whiffy tho Jun 15 13:19:29 counter Jun 15 13:19:30 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 55%) P1: 3 days 11:20:13 (3.472±3.5 days) (1475;228) Jun 15 13:22:35 Openmoko: The story of the neverending three days Jun 15 13:22:51 counter Jun 15 13:22:52 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 55%) P1: 3 days 11:18:32 (3.471±3.5 days) (1476;228) Jun 15 13:29:50 zecke, ping Jun 15 13:47:03 Sup3rkiddo: pong Jun 15 13:48:28 zecke, hi, i presume you are working on python for openmoko as well Jun 15 13:50:08 Sup3rkiddo: sort of, this stupid gforge didn't allow me to import the code though Jun 15 13:51:12 zecke, cool...i am from the python-openmoko team..i guess it would make sense to know what each other is doing to prevent duplication of work..what are you working on now..btw? Jun 15 13:51:49 Sup3rkiddo: ~three month ago I created rather complete bindings Jun 15 13:52:06 zecke, yup, i know :) Jun 15 13:52:10 zecke: The gforge problem still persist? Is there a bugrport for it? Jun 15 13:52:45 stefan_schmidt: let me check Jun 15 13:55:09 stefan_schmidt: a) Developer Subversion Access via SSH is unlikely to work Jun 15 13:56:30 stefan_schmidt: svn: Can't create directory '/svnroot/bindings/db/transactions/0-1.txn': Permission denied Jun 15 13:56:59 zecke: *sigh* Jun 15 13:57:14 zecke: Please fill a bug so I can pester our admin more often Jun 15 13:57:14 stefan_schmidt: point me to the bugtracker url (which I'm currently searching) Jun 15 13:57:23 ttp://bugzilla.openmoko.org Jun 15 13:57:37 stefan_schmidt: ah okay Jun 15 13:58:21 BTW: http://bugzilla.openmoko.org/cgi-bin/bugzilla/enter_bug.cgi has a spelling error on 'bugzilla' Jun 15 14:00:12 ah, site infrastructure. No idea who has access to this. Jun 15 14:00:33 stefan_schmidt: #610 has the rather short error text Jun 15 14:00:58 * * OM Bug 610 has been created by freyther(AT)yahoo.com Jun 15 14:00:59 * * Unable to do initial commit to the openmoko-bindings project Jun 15 14:01:01 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=610 Jun 15 14:01:20 zecke: Should be fine. He should try to reproduce it and ask for more info if needed Jun 15 14:03:36 back to work Jun 15 14:04:18 zecke, i have just wrapping up libgsmd...i couldnt access the bindings from the link you had given in the mailing lists Jun 15 14:04:24 *started Jun 15 14:18:56 (script) wiki RecentChanges: [[User:Minime/sketch]] Jun 15 14:30:17 Sup3rkiddo: http://page.mi.fu-berlin.de/~freyther/python-openmoko-0.3.tar.bz2 should just work Jun 15 14:30:52 zecke, thanks :), i will take a look Jun 15 14:31:23 Sup3rkiddo: and there is a rather good article on IBM developerworks from ross about wrapping gobjects in python Jun 15 14:31:50 zecke, yup i will go into that as well... Jun 15 14:34:50 zecke, fyi, i am using distutils at my side, and am wrapping up libgsmd directly.. Jun 15 14:35:17 Sup3rkiddo: why would one do that? Jun 15 14:35:38 zecke, you mean libgsmd? Jun 15 14:36:23 Sup3rkiddo: I have not looked at mickeyl's GObject wrapper for libgsmd but I assume it is more convient to use that instead of the 'raw' interface Jun 15 14:37:21 Wouldn't it make things more consistent as well? Jun 15 14:37:30 (using the gobject wrapper) Jun 15 14:37:41 Vegar: it would Jun 15 14:37:45 zecke, really?...gee. i am really new to this...will take a look at that Jun 15 14:40:17 zecke, ok then ..i will take a look at that and rewrite the stuff using the gobject wrapper Jun 15 14:40:55 Sup3rkiddo: as of distutils I have never really used it to comment on it Jun 15 14:41:42 zecke, na..actually i found it easy..thats why.. Jun 15 14:41:50 i am used it Jun 15 14:41:56 *-am Jun 15 14:42:54 ~seen sean Jun 15 14:42:57 sean was last seen on IRC in channel #debian, 65d 5h 31m 13s ago, saying: 'i demand an ircop to assist me'. Jun 15 14:43:46 not that sean I guess ;-) (IIRC he did visit here at some point) Jun 15 14:47:46 ~seen mosko Jun 15 14:47:48 stefan_schmidt: i haven't seen 'mosko' Jun 15 14:48:03 apt: You should know better. :) Jun 15 14:53:38 stefan_schmidt: I think we was using sean nick or close to it some months ago from some cyber cafe in china or something like that Jun 15 14:54:30 aloril: his normal nick is mosko. Jun 15 14:55:19 2007-02-04 15:15:23 quite a bit of people sitting and not much talking. How's everyone doing? [I'm sitting behind a painfully slow firewallin a teashop in Shanghai] Jun 15 14:56:39 aloril: So he used this one in a cybercafe. But I don't get why we discuss this. :) Jun 15 14:57:48 stefan_schmidt: because somebody asked about sean some time ago and then I did ~seen sean ;-) Jun 15 14:58:04 anyway, feels like ages ago looking at that discussion ;-) Jun 15 15:15:33 * aloril thinks its easier to write code than English ;-) Jun 15 15:36:42 /win 6 Jun 15 15:36:48 whoops. Jun 15 15:37:12 /loose 5 Jun 15 15:37:35 im still up by one. yea. :) Jun 15 15:37:43 hehe Jun 15 15:48:36 http://gnumonks.org/~laforge/weblog/2007/06/15#20070615-gta01-factory-trip Jun 15 15:49:06 yippie Jun 15 15:49:57 (script) planet: Harald "LaF0rge" Welte: Visiting FIC's factory for GTA01 mass production in Suzhou http://gnumonks.org/~laforge/weblog/2007/06/15#20070615-gta01-factory-trip Jun 15 15:50:12 now that's an enlighting posting Jun 15 15:50:20 especially compared to my secretive one :) Jun 15 15:52:01 DukeOfURL: awesome. thanks for the link! Jun 15 15:52:40 I wish for more information concerning FCC approval. Jun 15 15:52:53 hmm Jun 15 15:52:55 * mickey|fic misses CIAbot Jun 15 15:53:46 DukeOfURL: there has been stuff earlier saying it's been applied for Jun 15 15:54:11 any indication of teim frames? Jun 15 15:54:11 yay! Jun 15 15:54:16 time frames? Jun 15 15:55:36 * aloril wonders how long it does take before web-shop is up and running Jun 15 15:56:39 * SpeedEvil sets openmoko.com on 5s reload Jun 15 15:57:09 SpeedEvil: now that is evil :) Jun 15 15:58:50 SpeedEvil: hey, ditch that pull stuff. we're going to give you an irq when it's ready ;) Jun 15 15:59:07 busy waiting is so yesterday Jun 15 15:59:24 Naah, in rality, I'm just lurking on IRC. Jun 15 15:59:59 Are these all black/silver ones? Jun 15 16:00:08 good question. i have no idea Jun 15 16:00:11 if I'm reading that right, does that mean effectively there will be ~1000 units in total (eventually)? Jun 15 16:00:13 i'll ask Jun 15 16:00:43 so_solid_moo: yes, that's more or less the sum for the next couple of weeks Jun 15 16:00:58 mickey|fic: great. I think that's a pretty good number. Jun 15 16:01:19 ya, should be ok for the first wave of demand. lets see Jun 15 16:01:44 I hope not. Jun 15 16:01:47 :) Jun 15 16:01:55 right Jun 15 16:01:55 I think you all deserve a round of applause, btw, because I'm sure it must be quite disheartening in many ways for things to drag on so long. Jun 15 16:02:24 I'm just hoping that the factory doesn't get hit by a flood/asteroid strike/herd of rampaging elephants. Jun 15 16:02:36 we're following winston churchill Jun 15 16:02:40 "If you are going through hell, keep going" Jun 15 16:02:46 :) Jun 15 16:02:50 SpeedEvil: don't jinx it! Jun 15 16:03:51 SpeedEvil, Suzhou is further in land than Shanghai Jun 15 16:04:09 Asteroids can hit anywhere! Jun 15 16:06:02 * aloril reduces planet poll time from 30min to 20min hoping its OK (and when changes happen they are announced here) Jun 15 16:09:44 openmoko.com is polled every 10 minutes at the moment Jun 15 16:09:53 gtg, cu Jun 15 16:10:18 hrw|gone: my cf8385 driver seess to work now, I can ping in and out, ssh in and out Jun 15 16:10:24 * aloril suspects those 492 units are going to sell fast and thus for some there will be another 2 week wait Jun 15 16:10:26 hrw|gone: I need TESTERS! :-) Jun 15 16:10:55 schurig does the Steve Jobs: TESTERS TESTERS TESTERS. Jun 15 16:10:59 mickey|fic: Will the store have 'spare parts' available? - cases, forex? Jun 15 16:11:02 or was it his assistant? Jun 15 16:11:16 mickey|fic: ? Jun 15 16:11:20 SpeedEvil: not planned. sounds like an idea though Jun 15 16:11:22 Though I can see screens being of use too. Jun 15 16:11:36 mickey|fic: I'm too dump, I don't get what you mean Jun 15 16:11:36 schurig: never seen the developer conference where he said "developers developers developers" ? Jun 15 16:11:47 mickey|fic: no, not really Jun 15 16:11:50 that's quite famous Jun 15 16:11:54 ok, nevermind then :D Jun 15 16:12:22 For example, I want to be able to hack on the case, without worrying about it being irreplacable. Jun 15 16:12:29 mickey|fic: months ago hrw got one cf8385 card for free because he wanted to write a driver for it ... didn't do anything, shame on him :-) Jun 15 16:12:43 Adding joysticks/buttons/connectors. Jun 15 16:12:46 mickey|fic: but because I know that he has such a card, I'd like to cross-test what I've made so far Jun 15 16:13:11 Has anyone tried 2G microSD on it? Jun 15 16:13:24 mickey|fic: my wife is coming, picking me up. Greetings from her! Jun 15 16:14:41 SpeedEvil: on the NEo? Jun 15 16:16:55 balrog-kun: Well - yes. Jun 15 16:17:45 SpeedEvil: yes, i have used a 2G microSD card Jun 15 16:17:59 :) Jun 15 16:18:14 I haven't seen any actual orderable 4G ones though. Jun 15 16:18:36 hi Jun 15 16:18:43 hi Jun 15 16:19:31 I wonder if Samsung's F700 will be able to do openmoko Jun 15 16:19:34 i haven't either Jun 15 16:19:57 SpeedEvil: me neither... but is it possible? 16Gb on a chip that small? Jun 15 16:21:00 i wanted a 256 or 512 MB one just to have some storage but the person who gave me the card decided it wasn't educated to buy someone a card below 2GB for birthday Jun 15 16:21:18 so 2GB is more than i need :) Jun 15 16:22:41 LuitvD: has any information been leaked about the F700's cpu by now? Jun 15 16:23:04 I haven't been able to find some info that's that specific... Jun 15 16:23:17 which is strange, because the devices are ready to be shipped Jun 15 16:23:58 though Samsung seems to have decided to wait some 5 weeks before releasing... Jun 15 16:24:36 tri: ping Jun 15 16:24:58 mickey|fic: pong Jun 15 16:25:42 hey tri. are you the guy behind exactcode.de ? Jun 15 16:26:47 mickey|fic: no - that's rene aka rxr - you may found him in #t2 Jun 15 16:27:04 ah, ok. sorry Jun 15 16:27:48 mickey|fic: np i'm just a t2 user :-) Jun 15 16:27:55 :) Jun 15 16:28:03 t2 is in a nutshell what? Jun 15 16:28:27 ah nm Jun 15 16:28:29 reading it atm. Jun 15 16:29:22 http://www.t2-project.org/handbook/html/t2-book.html is a nice starting point Jun 15 16:29:29 bloody hell, AVR linux? :| Jun 15 16:29:46 must be AVR-32 or not? Jun 15 16:44:25 drat, seems I can't update counter page, that server is way too loaded from other stuff (will try again) Jun 15 16:57:30 aloril - With what info did you intend to update? Jun 15 17:02:01 counter_msg http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) Jun 15 17:02:01 aloril changed counter prefix message to http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) Jun 15 17:02:16 Elrond: seems load passed, managed to update it Jun 15 17:05:28 combined probability until 2007-07-07 is now 90% (dunno if its too high, but ... ;-) Jun 15 17:08:58 zecke: ping Jun 15 17:09:05 zecke: komm mal nach opie.de :) Jun 15 17:16:43 aloril: too high. Jun 15 17:17:17 (it's a loooooooooooooooooooong way from a factory lab in China to the shiipping department. Jun 15 17:17:19 ) Jun 15 17:18:29 * mwester will give 90% probability to Sept 1st, 30% July 1st, and 60% to August 1st) Jun 15 17:21:59 mickey|fic: the "developer, developer, developer"-dude was steve ballmer, not jobs ;) Jun 15 17:22:08 * balrog-kun agrees with mwester Jun 15 17:23:15 miip: aah right Jun 15 17:23:23 one of those steves Jun 15 17:23:24 cough Jun 15 17:23:47 heh :) Jun 15 17:23:53 hehe Jun 15 17:24:18 Neither of whom should be confused with Steve Ballmer Jun 15 17:27:45 Balmer's "Developers" http://youtube.com/watch?v=NSIMeRtVebM Jun 15 17:30:40 the amount of sweat :D Jun 15 17:31:33 hehe Jun 15 17:39:30 anyone have problems building openmoko-libs today. This is my pastebin: pastebin.no/2058 Jun 15 17:39:58 goxboxlive - Which svn-rev? Jun 15 17:40:03 i am building from the oe.dev , not the mokobuild Jun 15 17:40:12 Elrond, let me check Jun 15 17:40:43 Huh, oe.dev? Jun 15 17:40:56 rev 2261 Jun 15 17:41:07 DEPENDS += libxosd Jun 15 17:41:10 g'night Jun 15 17:42:14 goxboxlive - builds fine for me, by hand. See mickey|zzZZzz, you need libxosd. Jun 15 17:42:24 yes i added it now, thx Jun 15 17:47:01 mwester: oh, in Taiwan from production to shipping was faster Jun 15 17:54:00 But wasn't that only a small handful of devices? With no "official" packaging, such as the product box, and cardboard/plastic containers to keep all safe? Jun 15 17:57:24 * mwester is envisioning another wait, as the finished Neo's wait for a turn at the packaging line, to be united with the (hopefully already-in-stock!!! Does anyone know?) batteries, charger, guitar pick :-) etc, in a box suitable for international shipping, whilst the Moko team works on the legalities of where it can be shipped, and where not... etc. Jun 15 17:58:23 Which countries would not be allowed to ship it to? Jun 15 18:00:25 hey, harald's posting sounds really promising Jun 15 18:00:29 Perhaps ship to is not the concern -- but AFAIK it is not yet FCC approved, so cannot be sold here -- is that not right? Jun 15 18:01:13 sounds really as if the neos are avail in july Jun 15 18:01:31 good news that Jun 15 18:02:38 mwester - I have no idea on fcc (or CE) Jun 15 18:03:46 it's ashame that the car holder is so lousily overpriced Jun 15 18:04:35 I'll rather get a generic one from ebay for ten bucks Jun 15 18:05:47 evening Jun 15 18:06:00 Hi Rince. Jun 15 18:06:06 tihi: Jun 15 18:06:07 bin% ldd openmoko-messages | wc -l Jun 15 18:06:07 79 Jun 15 18:06:16 That's embedded? ;o) Jun 15 18:06:32 (OKay, I admit, this is build on my host. ;o) ) Jun 15 18:07:52 ** ERROR **: Not enough memory to set up DBusConnection for use with GLib Jun 15 18:07:58 *lol* ;-) Jun 15 18:08:14 ouch Jun 15 18:08:19 have you cut down to 640K? Jun 15 18:09:14 gamin - That box has 256MB. ;-) Jun 15 18:09:43 MB or KB ? Jun 15 18:09:46 *lol* Jun 15 18:09:59 MB :-) Jun 15 18:10:40 is the dialer app stable now? i.e. is the phone usable without cat-ting to /proc? Jun 15 18:10:56 echo'ing to /proc that is Jun 15 18:11:21 echo'ing to /proc?! Jun 15 18:11:52 echo 1 > /proc/foo/bar/neo/gsm/up/now - whatever Jun 15 18:12:27 harald did so in the june video Jun 15 18:12:54 but then he will probably do even if there is a working ui ;) Jun 15 18:13:48 You mean /sys, I guess (also I never fully understood, why we need /foo, /bar, and /frob every new kernel major release.) Jun 15 18:14:45 Elrond: so do you use your neo on a daily basis? Jun 15 18:15:40 gamin - As phone? No. ;-) Jun 15 18:15:44 seems not to stable either, the n800 >:-) Jun 15 18:15:50 s/to/too/ Jun 15 18:15:50 gamin meant: seems not too stable either, the n800 >:-) Jun 15 18:16:07 so just as a play tool? Jun 15 18:18:00 gamin - I'm mostly working on the gps. Jun 15 18:18:15 oh, sure. you are osm. Jun 15 18:18:28 how could I possibly forget that ;) Jun 15 18:18:46 Huh? Jun 15 18:18:55 aren't you? Jun 15 18:19:02 A bit, yes. Jun 15 18:19:06 see Jun 15 18:19:07 But how did you guess? Jun 15 18:19:16 same nick over there Jun 15 18:19:21 Ahh ;) Jun 15 18:19:28 I haven't been on #osm in a while. Jun 15 18:19:37 neither was I Jun 15 18:19:54 but planning for the return Jun 15 18:20:04 nevertheless: you work on the gpsd? Jun 15 18:20:09 or on app level? Jun 15 18:20:37 I'm hacking on the hammerhead protocol stuff. Jun 15 18:20:45 Elrond: any news on that already? Jun 15 18:21:05 pH5 - I'm currently addicted by "simple build of OM-2007 on host" ;o) Jun 15 18:21:13 so you are NDA'd? or is there some open stuff / you are doing reverse engineering? Jun 15 18:21:15 ... soon at a wiki near you. ;o) Jun 15 18:21:43 Elrond: you go to guadec? Jun 15 18:21:48 gamin - I'm doing reverse engeenering. I haven't seen any NDA'd material, nor did I sign any NDA. Jun 15 18:22:04 guadec? Jun 15 18:22:19 I read somewhere a remark that they want to get rid of the hammerhead anyway Jun 15 18:22:42 guadec: gnome conferenc in birmingham with a mapping party of osm Jun 15 18:22:56 Ahh. No. Too far away. Jun 15 18:23:07 Where did you read this? Jun 15 18:23:30 Hmm... if it's true, I wonder, how much sense it makes to reverse engeener this thing anyway. Jun 15 18:24:04 Elrond: hmm, probably I heard it in the video. or harald or sean wrote it somewhere. maybe on the slides about gta02 Jun 15 18:24:32 Elrond: it think it still makes sense to reverse engineer it Jun 15 18:24:35 *i Jun 15 18:25:03 the same chip is in the tomtoms and other devices, and it would be a first GPS chip with opensource protocol then Jun 15 18:25:21 SpeedEvil believes it is some kind of very low-level sat data... Jun 15 18:25:47 balrog-kun: but the sirf's are documented? Jun 15 18:25:49 and i bet other models from TI and maybe even other manufacturers use similar protocols Jun 15 18:25:53 There are possible benefits to being able to talk very low level. Jun 15 18:26:08 hey SpeedEvil Jun 15 18:26:08 gamin: they are? Jun 15 18:26:31 balrog-kun: search google for sirf binary protocol Jun 15 18:26:33 For example, it could buy - with a GPRS 1K or so data packet from a server with local ionospheric corrections - 1m or better position for a little while. Jun 15 18:26:47 counter Jun 15 18:26:47 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: 3 days 08:46:34 (3.366±3.4 days) (1477;228) Jun 15 18:26:55 The SiRF is much more expensive AIUI. Jun 15 18:27:22 http://wiki.openmoko.org/wiki/Hardware:AGPS see the end Jun 15 18:27:27 gamin: humm, indeed Jun 15 18:27:49 Look for DGPS Jun 15 18:28:04 SpeedEvil: that is still not correct ;-p Jun 15 18:28:15 It's nice in that you can do some potentially really quite clever stuff. Jun 15 18:28:22 In what way. Jun 15 18:28:35 With the CPU talking directly to the low-level hardware. Jun 15 18:28:48 gamin: Can you point at an incorrectness. Jun 15 18:28:49 the output of the chip is a lot higher level than you said Jun 15 18:28:59 gamin: based on what information? Jun 15 18:29:07 SpeedEvil: give me a minute Jun 15 18:33:32 gamin: I'm basing my assertions off the published product brief and tech reports available on the Global Locate website, and the knowledge I gained when making an actual hardware receiver. Also recent research. Jun 15 18:34:44 (A really simple one, that only picked up one SV at once.) Jun 15 18:37:02 SpeedEvil: you said the hammerhead spits out: "`I am recieving a `1` from satellite with PRN code 12, at a doppler shift of 1.03Km/s towards us, with a time offset of 1283.4ns" Jun 15 18:37:39 thus the CPU of the neo would need to do all the relativistic stuff corrections, apart from lots of other stuff... Jun 15 18:37:43 Right - that's simplified - the range and velocity are combined. Jun 15 18:37:58 It's not an actual range or velocity. Jun 15 18:38:29 that is a hell of a lot of floating point stuff... Jun 15 18:38:38 It's a psuedorange, which is a relative doppler shift between the recieved signal, and a local clock, and the offset to that local clock. Jun 15 18:39:10 no to spam the IRC here a bit: Jun 15 18:39:13 The GL16000 chip architecture is unlike any other GPS Jun 15 18:39:13 receiver. In a traditional GPS architecture, the Jun 15 18:39:13 microprocessor has significant responsibility. The Jun 15 18:39:13 hardware typically ends in a set of registers containing Jun 15 18:39:13 correlation results for the past millisecond. The CPU must Jun 15 18:39:14 And it's not an offset between the 'times' it's an offset between the local clock, and the data bits. Jun 15 18:39:14 quickly read these registers. In a typical scheme, the CPU Jun 15 18:39:16 services interrupts at a steady 2 kHz rate, constantly Jun 15 18:39:18 checking all the hardware channels and processing any Jun 15 18:39:20 waiting results. In this scheme, the CPU load is driven by Jun 15 18:39:22 the need to read, store, and process correlation results in Jun 15 18:39:24 real time, and rises proportionally with the number of Jun 15 18:39:26 correlators in the design. With thousands of correlators, Jun 15 18:39:28 an extremely high performance CPU would be required. Jun 15 18:39:30 Furthermore, the interrupt-driven processing must run at Jun 15 18:39:32 high priority, making for difficult integration with other Jun 15 18:39:34 software tasks. For these reasons, the conventional Jun 15 18:39:36 architecture has characteristics that make it problematical Jun 15 18:39:38 for GPS software functions to be successfully hosted in an Jun 15 18:39:40 already busy wireless CPU, especially if correlator count Jun 15 18:39:42 is to be increased. Jun 15 18:39:44 sorry for that, it is from globallocates docs... Jun 15 18:39:45 I know, I've read this. Jun 15 18:39:52 It has the integrators internal. Jun 15 18:39:55 and it contradicts what you say... Jun 15 18:40:03 No, it doesn't. Jun 15 18:40:20 It has variable period integrators internally. Jun 15 18:40:36 You set these up, and then later it reports the result of the correlation it's just done. Jun 15 18:40:38 well, if it has the integrators internal, than it wont spit out"`I am recieving a `1` from satellite with PRN code 12, at a doppler shift of 1.03Km/s towards us, with a time offset of 1283.4ns" Jun 15 18:40:50 No, it won't. That was a simplification. Jun 15 18:41:07 The actual output would be more along the lines of: Jun 15 18:41:10 ah, okay, then we misunderstood each other Jun 15 18:42:12 I am recieving a '1' from a satellite with a PRN code of 12, at a doppler shift of 1Km/s against the local clock, with a time offset of 1233.4 nanoseconds from the clock. Jun 15 18:42:33 And the local clock may drift a bit, this is understood, and is taken into account in calculations. Jun 15 18:42:58 The important thing is the relative dopplers and offsets between recieved satellites. Jun 15 18:43:20 Unless you only care about time to within a few millieseconds, in which case you only need one. Jun 15 18:43:30 SpeedEvil: you can write stuff to the channels and read it back, so I guess it doesn't only return a '1' but has some kind of buffer per channel, too. Jun 15 18:43:49 Yeah - it's probably not going to be that simple. Jun 15 18:44:03 It'll probably be a 2-3 bit value Jun 15 18:44:10 rather than '1' or '0' Jun 15 18:44:16 and maybe I and Q values too. Jun 15 18:44:57 I still believe it is a lot higher level than that. Jun 15 18:45:27 Look at the schematic diagram of the hammerhead chip. Jun 15 18:47:05 and look at the protocol, from the amount of interaction, it certainly doesn't look very high-level. Jun 15 18:47:12 http://globallocate.com/Newest_Illos/SEMI_HAMMERHEAD_dia_AA.jpg Jun 15 18:48:31 that does support my position more than yours? what else is the postprocessor good for? Jun 15 18:48:44 Also - look at the product brief. Jun 15 18:49:12 It _only_ lists correlators in the features. Jun 15 18:49:24 It does not list any more 'cooked' features. Jun 15 18:49:45 Exactly how cooked are you saying the data it outputs is? Jun 15 18:50:12 similar to the sirf binary Jun 15 18:50:39 ..protocol Jun 15 18:51:29 'Both the A-GPS control software and LBS (location based services) application are executed on the host CPU of the cellular processor, so additional external CPU or memory is not required' Jun 15 18:51:47 so what? Jun 15 18:52:13 the agps-control is just the stuff shifting the ephemeris into the chip Jun 15 18:52:32 gamin - It's not as cooked as the sirf. One has to setup the channels from the CPU, etc. Jun 15 18:52:39 gamin - And try PRNs, etc. Jun 15 18:52:59 gamin - My first goal is "search satelites, that are receivable". Jun 15 18:53:48 than actually they (GL) are lying: the will need quite a lot of CPU... Jun 15 18:54:04 No, it doesn't. Jun 15 18:54:21 You readout the values from the integrators, probably already thresholded. Jun 15 18:54:30 This is not time-critical, nor is it hard. Jun 15 18:54:35 ls Jun 15 18:54:42 My garmin GPS-12 can do it with a 386/16 Jun 15 18:54:56 And display a moving track at the same time. Jun 15 18:55:33 Working out satelite orbits is near-trivial. Jun 15 18:55:40 The code is really quite easy. Jun 15 18:55:49 And it's not CPU intensive. Jun 15 18:56:08 Nor is converting the received psuedoranges into a position. The datarates are quite low. Jun 15 18:57:19 The algorithms are annoying to program, but not actually computationally intensive. Jun 15 18:58:31 aren't lots of the alogrithms already out there? seemed like a quite active community to me... Jun 15 18:58:41 and all scientists Jun 15 18:59:07 The algorithms for going from psuedorange - position are already out there. Jun 15 18:59:29 There is an opensource package with all sorts of utilities like this in is available. Jun 15 18:59:34 I forget the name. Jun 15 19:00:46 gpstk? Jun 15 19:01:04 Sounds possible. Jun 15 19:01:05 http://www.gpstk.org Jun 15 19:02:25 SpeedEvil - garmin gps-12 needs a 386?! Jun 15 19:02:52 I'm just saying what I found when I opened it. Jun 15 19:03:13 It's now re-sealed, with duct-tape. Jun 15 19:03:40 I was trying to take it apart, to see if I could fix the dead column in the graphic LCD - I couldn't. Jun 15 19:03:59 386ex16 - it was not expensive at the time. Jun 15 19:04:23 I don't know that it actually runs at full speed though. Jun 15 19:05:04 Fun. Jun 15 19:05:11 what is the price of a GL hammerhead compared to a sirfII? Jun 15 19:05:14 Doesn't a 386 take a lot of power? Jun 15 19:05:33 (or is the gps-12 anyway for wired using". Jun 15 19:05:45 SpeedEvil says it doesn't :) Jun 15 19:05:51 Elrond: 4 AA batteries - it takes some 100mA at 6V Jun 15 19:06:09 0.6W ... Jun 15 19:06:10 Paypal me $1, and I'll even put up some pictures! Jun 15 19:06:46 may i fax you the 1$ instead? :) Jun 15 19:07:24 I have no idea, what the GL HH costs... Jun 15 19:07:33 I know, what a sirfIII costs... Jun 15 19:07:51 yes? Jun 15 19:08:14 I think they are a couple of $$ Jun 15 19:08:23 http://spatialnews.geocomm.com/features/gpsreview/gps.htmlhttp://spatialnews.geocomm.com/features/gpsreview/gps.html Jun 15 19:08:28 http://spatialnews.geocomm.com/features/gpsreview/gps.html even Jun 15 19:08:39 reference to 386 in garmin GPS Jun 15 19:09:23 the cheap CF-GPS w/ sirf III are roughly 45$. Jun 15 19:09:57 so the chip may be 6-7$ ?? Jun 15 19:10:25 The hammerhead has additional plusses though. It's one chip. Jun 15 19:10:34 So may be smaller and easier to integrate Jun 15 19:10:36 mwester: for mass market and 30-300x more devices: yeah I guess it would take some time to go from factory to customer Jun 15 19:10:42 the sirf's are one chip too Jun 15 19:11:14 last time i read they where 5x5mm Jun 15 19:11:57 * aloril2 hopes they don't use any extra time for fancy packaging stuff for developers release: P0 packaging is just fine for me Jun 15 19:11:58 The sirfIII is quite large. Jun 15 19:12:05 I don't know the size of the hh though. Jun 15 19:13:43 oh well, the very latest hh is 3.7x3.5mm Jun 15 19:14:15 and much better reception than sirf Jun 15 19:14:22 dare you if you sneeze in front of a bag of those... Jun 15 19:14:44 Hopscotch: both claim -159dbm Jun 15 19:15:12 and in test the HTC devices didn't perform that well. Jun 15 19:15:36 yeah, but tomtom uses sirf3 and hh Jun 15 19:15:41 and the hh work much better Jun 15 19:15:44 gamin: it's difficult to compare this, because different devices have different antennas Jun 15 19:17:51 Hopscotch: any pointers as to what exactly means "much better" Jun 15 19:18:09 difficult to say as i haven't any measurement equiment here Jun 15 19:18:22 but the sirf3 only works when i have a streight view to the sky Jun 15 19:18:39 the hh works inside my flat which happens to be in the uppermost floor Jun 15 19:18:49 the roof is flat Jun 15 19:19:45 i guess the difference is a few db - not much but enough Jun 15 19:20:45 I have only a cheap CF-mouse with a SiRFIII and it works indoors with no view to the sky. and not only when in tracking mode but even when cold starting. but takes then a good 20minutes to first fix - as opposed to less than two when outside Jun 15 19:22:00 1st floor of five storey brick house Jun 15 19:22:31 gamin: this is all quite irrelevant at the moment. Jun 15 19:22:39 The GPS in the device is the hammerhead. Jun 15 19:22:51 There have been no indications from FIC that this will change. Jun 15 19:23:07 And the GPS in it seems to work - even with the supplied software. Jun 15 19:23:14 You may also want to read the protocol dumps. Jun 15 19:23:35 It 'feels' much more like a 'call and response' protocol, rather than Jun 15 19:23:47 an init, and get back position. Jun 15 19:24:17 SpeedEvil: I think it was Harald to make the remark that they try to change the GPS chip. Jun 15 19:24:55 I guess in the video where he talk about the hardware changes from gta01 to gta02 Jun 15 19:25:15 SpeedEvil: the hammerhead is dumb stupid Jun 15 19:25:28 there are only correlators and the simple serial interface in it Jun 15 19:26:11 Hopscotch - No thresholding or stuff? Jun 15 19:26:35 Hopscotch: that's not quite correct. There is also reference in the docs to variable period integrators. Jun 15 19:26:40 Hopscotch - That would mean 2046 things to readout to the CPU per channel... which is a lot. Jun 15 19:26:51 yes, ture Jun 15 19:26:52 true Jun 15 19:26:53 And which is also not visible in the output from the chip. Jun 15 19:27:18 my gps knowledge is very limited Jun 15 19:28:21 There are a couple of nice tutorials in as much depth as you could want - almost - at the beginning of http://wiki.openmoko.org/wiki/Hammerhead_protocol Jun 15 19:30:07 whow - lots of new stuff Jun 15 19:30:26 The official protocol specs are not horrible - at least the first versions. Jun 15 19:30:44 The latest protocol docs seem to have been written to obscure things. Jun 15 19:30:55 Huh? Jun 15 19:31:39 The official GPS protocol docs. Jun 15 19:34:35 Ahh. A friend downloaded them and was scared away. Jun 15 19:36:32 * SpeedEvil contemplates text entry cases. A 4chan browser would be really quite simple. You'd need about 6 buttons. Jun 15 19:46:49 mwester: especially mass market amout at chinese factory -> brick & mortar store for end user could take a while Jun 15 19:47:26 mwester: anyway, I'll adjust probabilities down tomorrow Jun 15 19:47:28 SpeedEvil: i'm actually trying out hexinput :) Jun 15 19:57:20 guaqua: I was more meaning that you'd only need 6 phrases. Jun 15 20:06:57 (script) wiki RecentChanges: [[User:Minime]] [[User:Minime/sketch]] Jun 15 20:09:08 SpeedEvil: that too :) Jun 15 20:14:21 SpeedEvil: so does it make sense that the hh knows about specific sats? Jun 15 20:14:38 SpeedEvil: or does it make more sense that it knows about PRNs? Jun 15 20:19:59 yay, flashing through u-boot's DFU is much faster now Jun 15 20:26:26 has anyone tried making the phone resume on incoming calls? Jun 15 20:26:44 if it's suspended at that time Jun 15 20:29:20 balrog-kun: I once did that for a different device. The GSM module had a serial port, and I programmed the RI(ng) GPIO into wakeup mode. Worked quite nice. Jun 15 20:31:30 schurig: i know on the iPaq h5400 it's done just by setting the UART as an wakeup event source, you just need to make sure that the modem doesn't report too often Jun 15 20:31:49 for example turn off the signal quality reporting Jun 15 20:32:06 balrog-kun: on my device the GSM module was hooked to one of the three builtin UARTs of the PXA255 Jun 15 20:32:10 s/was/is/ Jun 15 20:32:11 schurig meant: balrog-kun: on my device the GSM module is hooked to one of the three builtin UARTs of the PXA255 Jun 15 20:32:32 i see Jun 15 20:32:49 in h5400 it's hooked to one of the three OMAP UARTs Jun 15 20:33:46 balrog-kun: err, h5400 doesn't have OMAP :) Jun 15 20:33:54 it's PXA255:) Jun 15 20:34:15 hmm, maybe it's not 5400 that i'm talking about then, lemme check Jun 15 20:34:41 I have not seen any GSM sleeve yet, so I don't know... Jun 15 20:35:13 mmp: oh, right, it was the h6300 Jun 15 20:36:18 :-) just a note for record Jun 15 20:36:28 * mmp looks at his preeeciouus h5550 :) Jun 15 20:40:16 * balrog-kun doesn't have an iPaq but has a device with the same audio codec chip as the h6300 so i worked together with the h6300 port author Jun 15 20:42:22 :) Jun 15 20:45:41 ~counter Jun 15 20:45:43 or Jun 15 20:45:44 counter Jun 15 20:45:45 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: 3 days 07:37:05 (3.317±3.3 days) (1478;228) Jun 15 21:00:37 hmm, actually the noise that qemu-neo1973 makes in some way resembles what the real device emits, now that i know how it sounds Jun 15 21:19:24 hi pingu^ Jun 15 21:21:36 * Elrond loves hardware wahtchpoints. ;-) Jun 15 21:25:49 counter Jun 15 21:25:49 http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter (probability guess: 65%) P1: 3 days 07:17:03 (3.304±3.3 days) (1479;229) Jun 15 21:25:55 Still 3 days :] Jun 15 22:14:58 * * OM Bug 611 has been created by pavel(AT)ucw.cz Jun 15 22:14:59 * * python-gtk as .ipk would be nice Jun 15 22:15:01 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=611 Jun 15 22:15:09 * * OM Bug 612 has been created by  Jun 15 22:15:11 * * kui globals make openmoko-messages crash Jun 15 22:15:13 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=612 Jun 15 22:15:59 Bug!! :O Jun 15 22:16:30 Bug fixed in minutes!! :D Jun 15 22:16:34 * nox-Hand hugs OpenMoko Jun 15 22:47:52 rwhitby now has a 16MB flash, 256MB ram, "ObeseSlug" http://www.nslu2-linux.org/wiki/HowTo/ObeseSlug Jun 15 22:59:35 rwhitby - BTW: I have some "simple build parts of openmoko for host" shell thingy. it might be an addition to MokoMakefile. ;-) Jun 15 23:00:28 sure Jun 15 23:00:40 URL? Jun 15 23:01:39 WARNING: It's not yet 100% tested Jun 15 23:02:14 Hmm, anyone a file pastebin? My webservers are all a bit... strange today. Jun 15 23:06:20 * LuitvD initiates bed-going sequence Jun 15 23:06:26 bye all Jun 15 23:10:41 rwhitby - See query. Jun 15 23:16:37 * zecke heads home Jun 15 23:27:35 rwhitby - *ping* Jun 15 23:29:45 Elrond - having network problems here. Jun 15 23:30:10 rwhitby-n800 - Yuck.... Jun 15 23:37:13 too much RAM for the nslu :) Jun 15 23:38:03 can never have too much ram in an nslu2 :-) Jun 15 23:38:12 (since swapping to usb disk is so slow) Jun 15 23:38:15 5G? Jun 15 23:38:24 damn it just started to rain *trapped in the office* Jun 15 23:38:28 SpeedEvil: perfect for tmpfs ... Jun 15 23:38:40 Won't address it though. Jun 15 23:38:56 zecke: Just sleep two hours fetch some chinese food and hack again. ;) Jun 15 23:39:10 Elrond: can you send me that again? Jun 15 23:40:08 stefan_schmidt: well, that is the plan besides making chinese food myself and watching futurama while hacking Jun 15 23:40:33 zecke: heh Jun 15 23:44:00 * stefan_schmidt heads to bed. night folks Jun 15 23:44:29 cya Jun 15 23:55:15 Night stefan_schmidt Jun 15 23:58:04 N8 Jun 16 00:00:39 the rain almost stopped, I'm running home Jun 16 00:00:39 cya Jun 16 00:01:47 * mwester wonders if zecke is concerned about melting in the rain? Jun 16 00:51:18 i just read up on the latest posts on planet.openmoko.org and this looks really good! Jun 16 00:58:40 i guess it will stop the flow of complaints about vapourware on the community ML for a while :-) Jun 16 00:59:17 aevin: let's hope so Jun 16 01:08:56 Will OpenMoko be using a WebKit based browser for web browsing? Jun 16 01:09:42 I noticed zecke's post on Planet OpenMoko, and I think that I have seen him around #webkit. Jun 16 01:15:30 dkirker: that's the plan I believe, yep Jun 16 01:15:35 cool! Jun 16 02:49:49 svn: Unable to find repository location for 'http://svn.openmoko.org/trunk/src/target/u-boot/patches' in revision 11 Jun 16 02:50:19 is the uboot url wrong? **** ENDING LOGGING AT Sat Jun 16 02:59:56 2007