**** BEGIN LOGGING AT Sun May 06 02:59:56 2007 May 06 05:36:28 (script) wiki RecentChanges: [[Wishlist:Wikipedia_Mirror]] May 06 07:36:28 (script) wiki RecentChanges: [[Wish_List_-_Hardware]] [[Wish_List]] May 06 07:38:17 good morning May 06 07:40:33 good morning! May 06 08:11:26 hi May 06 08:33:02 Hi May 06 08:33:03 counter May 06 08:33:04 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, a day 15:26:54 (8.644 +-3.5 days) (1098;195) May 06 09:23:32 openmoko: 03laforge * r1915 10/trunk/src/target/u-boot/patches/uboot-s3c2440.patch: May 06 09:23:32 openmoko: * Fix S3C2440 support in the MMC/SD driver May 06 09:23:32 openmoko: * This also fixes 2410 SD card support problems (http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=164) caused by MMC style clocking rather than SD style clocking May 06 09:47:43 LaF0rge: Does this commit solve the "SD not always found / accessible after reboot" problem? May 06 10:28:04 balrog-kun: so there is no USB driver for qemu? May 06 10:28:08 s/no/now May 06 10:30:22 Is it possible to buy an openmoko yet for development? And are there any requirments you need to fulfill to get one early? May 06 10:30:22 Mukunda: 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 06 10:31:22 Neat. May 06 10:31:41 Mukunda: Phase 1 counter: http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, a day 15:26:54 (8.644 +-3.5 days) (1098;195) May 06 10:34:45 If I want to update my openmoko to latest software... is it enough to use dfu-util to update kernel and rootfilesystem? May 06 10:34:45 pavlem: 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 06 10:34:49 Mukunda, phase 1 means everybody who wants can buy a device. its for developers and testers, but everyone can buy May 06 10:35:06 pavlem: yes, that should do May 06 10:35:24 thomasg_: yes I see, hmm but a new version with faster cpu will be out for the consumer release... And the price of these is a bit.. :-\ May 06 10:35:46 pavelm: i didn't have a chance to test uboot on Bv3 lately. I'll report when i did May 06 10:35:48 hm, imho the price is ok May 06 10:35:56 I agree May 06 10:36:10 but if you can wait for july you probably should wait May 06 10:36:19 thomasg_: the price is great... But just I don't want to buy two of them, as I'd like to get the version with the faster cpu... May 06 10:36:34 mickeyl: But given that I do not hack the uboot, I should not brick my hw, right? May 06 10:36:35 Fair enough May 06 10:36:47 yes, but if you buy p1+ in summer and bought p1 before youll get the p1+ cheaper May 06 10:36:50 Hm July? Sorry, I haven't been following the officially mailing list, just checked wikipedia which said consumer release was in Sept. May 06 10:36:51 (I do not want to set up JTAG). May 06 10:37:11 but nobody knows how much discount there will be. I think they will tell us at release of p1 May 06 10:39:07 thomasg_: sorry for asking stuff I'm sure I could find out for myself, but is p1+ the version that will be released for consumers. ie: the one with the faster cpu? May 06 10:41:47 yes. p1+ is the supposed hardware refresh that's on sale in autumn May 06 10:43:38 Hm, guess I'll see how much the discount will be. May 06 10:45:41 Moo^^: kindof May 06 10:46:08 Moo^^: i need to add some commands to make it usable yet, and write some usage information May 06 10:49:12 Ah awesome, P1+ might have wifi... That'd really open up the ability for cheap voip hotspots. May 06 10:49:43 pavelm: you can try it in qemu May 06 10:51:27 magon: qemu is hard to set up... May 06 10:51:46 Mukunda, p1+ is the pre-consumer device. it features the same hardware afaik, but it will get released 2 months before May 06 10:51:50 I try to do ./dfu-util -a 3 -D ./uImage-2.6.21.1-moko10-r0_0_1883_0-fic-gta01.bin .. and it just hangs. May 06 10:52:37 and p1+ might not have wifi, it WILL have wifi, thats nearby for sure :) May 06 10:53:52 What is the right time to start flashing? Just after clean shutdown? from uboot menu? May 06 10:54:05 it only listens from within uboot menu May 06 10:54:35 IIRC May 06 10:55:09 I get More than one DFU capable USB device found, you might try `--list' and then disconnect all but one device May 06 10:55:28 supply the device you want via extra arguments May 06 10:55:32 to dfu-util May 06 10:55:39 dunno which params offhand May 06 10:55:47 Heh... strange, it thought my usb bluetooth can be upgraded..? May 06 10:55:48 Hmm, so it was impossible to have the gps and gsm parts to be free software supported... errm, meaning able to work without binary only firmware? May 06 10:56:39 yes. gsm is impossible because of the law, gps is impossible because of the manufacturer. at least the latter might change in further hardware... May 06 10:56:44 mickeyl: It works now, thanks. May 06 10:56:45 Mukunda: binary only firmware is not a problem, binary only driver is a problem May 06 10:57:18 mickeyl: why gsm is imposible due to law? May 06 10:57:36 well May 06 10:57:46 apart from the problem that there is no free GSM stack May 06 10:57:51 you won't get a certification for that May 06 10:57:54 You'd be able to do funny things on the network that the network owners wouldn't allow probably. May 06 10:58:16 certification to GSM spec ? May 06 10:58:46 You need CE certification at least in Europe. It covers more stuff than just the GSM May 06 10:58:56 mickeyl: Are you sure? I guess BSD-licensed stack would be easy to certify, you'd just not be able to hack it... May 06 10:58:57 yeah. you are not allowed to operate a device on the GSM network that isn't certified May 06 10:59:03 for example, things like "no little parts which children could eat" May 06 10:59:30 but CE is just for radio signal strength and freq May 06 10:59:32 pavelm: even if there was a stack... no community would be able to pay for the certification. May 06 10:59:53 besides... the general legislation doesn't _want_ an open GSM stack, but that's a different issue May 06 10:59:56 mickeyl: Agreed with that. (Plus, opensource stack you can't upgrade would suck, anyway). May 06 11:00:40 ok so no GSM open stack May 06 11:00:50 mickeyl: there will be money for it - i am sure of it May 06 11:00:52 Anyway... If wifi can get down, it'd be possible to subvert the old GSM networks, by setting up wireless voip hotspots. May 06 11:01:17 Some people (chaoscamp has a link) are basically building open-source gsm stack. They want to decode gsm, intercept messages, decipher calls, etc. May 06 11:01:35 yes, software controlled radio is very promising May 06 11:01:47 ouch ouch, got dfu-error 71 while trying to flash rootfs. May 06 11:02:17 pavelm: if you have still a very ancient uboot, then it might have the power-down-on-timeout bug May 06 11:02:21 is ./openmoko-devel-image-fic-gta01-20070501054014.rootfs.jffs2 right thing to flash ? May 06 11:02:28 yes May 06 11:02:32 aha, so I simply need to press aux button few times, right? May 06 11:02:36 yeah May 06 11:02:47 i would recommend upgrading uboot in that case May 06 11:02:47 or upgrade u-boot May 06 11:02:51 *nod* May 06 11:03:45 mickeyl: I'm scared. If I mess up u-boot upgrade, I have a brick. May 06 11:04:02 pavelm: you don't have a debug board??? May 06 11:04:10 mickeyl: I consider even kernel upgrade big deal. May 06 11:04:18 mickeyl: Yes, I do, but I do not know how to use it. May 06 11:04:21 hehe May 06 11:04:23 ok May 06 11:04:54 i'm afraid in this stage of development we will have a lot of upgrades to follow May 06 11:04:55 (My original plan was to leave flash intact, and just do complete install on SD card) May 06 11:05:14 yeah, booting from SD is good plan May 06 11:05:22 pavelm: unbricking neo is absolutely painless thanks to the devirginizer thingy =) May 06 11:06:15 coredump: I guess I could do it... I'd just prefer not to. I already bricked my zaurus once... to find that they have two levels of BIOS, but it took me a day to recover it. May 06 11:06:33 *nod* it takes me always a day to recover my Neo as well May 06 11:06:35 I'm just saying May 06 11:06:42 but at least it's possible May 06 11:06:49 considering that a whole lot of devices aren't May 06 11:06:53 (recoverable) May 06 11:07:02 My very first u-boot upgrade went wrong -> brick =D May 06 11:07:15 rumours are the next major new hardware will get NOR May 06 11:07:18 mickey: Yes, openmoko is better than average in this respect. May 06 11:07:19 :D May 06 11:07:43 mickeyl: next major=1.5? May 06 11:07:46 How does NOR help? ... aha, that was the case of zaurus, too. NOR flash with "deeper bios". May 06 11:07:50 I like the booting from sd idea, because I could install gentoo on sd and leave the openmoko inside untouched :) May 06 11:07:53 buz: yea May 06 11:07:57 pavelm: exactly May 06 11:08:03 NOR = safety net :) May 06 11:08:12 i hope 1.5 is released on time ;) May 06 11:08:16 heh May 06 11:08:19 mickeyl: Just write protect the uboot? May 06 11:08:54 pavelm: w/ an ro u-boot you couldn't ship bug-fixes for it May 06 11:09:02 Can someone check http://wiki.openmoko.org/wiki/Flashing_openmoko for obvious mistakes? May 06 11:09:19 coredump: yes, just make sure there are no bugs in uboot :-)_ May 06 11:09:27 hehe May 06 11:09:32 pavelm: newer systems should know names for partitions May 06 11:09:34 i.e. May 06 11:09:37 dfu-utils -a rootfs May 06 11:09:42 less error prone than "3" May 06 11:09:52 (5) May 06 11:10:09 pavelm: http://rafb.net/p/9xCGfu11.html <- that's known-working May 06 11:10:26 please guys, use names May 06 11:10:39 any idea how i can prevent xorg to open the ps2 port ? May 06 11:10:41 perhaps we change partitioning May 06 11:10:53 rob_w: hacking the source :/ May 06 11:11:03 rob_w: or removing the device nodce May 06 11:11:05 node, even May 06 11:11:12 mickeyl, odd enough May 06 11:11:13 ~lart xorg for unconditionally opening all input nodes May 06 11:11:14 * apt squeezes xorg till xorg turns blue like papa smurf for unconditionally opening all input nodes May 06 11:11:24 yeah May 06 11:11:29 coredump: well, my update seemed to go well, too. May 06 11:11:43 this is a bad habit May 06 11:12:15 coredump: Now I'm looking at oops in jffs_gcd_mtd4, then system hang in "configuring web". May 06 11:12:59 pavelm: you did erase the rootfs via u-boot before flashing right? May 06 11:13:07 ah yes May 06 11:13:12 pavelm: your ancient u-boot kinda requires that =) May 06 11:13:14 old uboot has problems in dfu-upgrade May 06 11:13:33 aha, ok, no, I did not do that. May 06 11:13:38 can only safely upgrade if the new image is larger than the old one May 06 11:13:42 bbaib May 06 11:14:34 ...should I add some zeros at the end of jffs2 file? :-) May 06 11:15:01 padding to MAX_ROOTFS_SIZE would work, yes May 06 11:15:17 what is max_rootfs_size? May 06 11:15:21 but seriously, upgrade u-boot =) The old one is bug-ridden May 06 11:15:37 pavelm: no idea. May 06 11:16:39 jffs2 is currently 27MB, will make it 32MB or so. May 06 11:17:22 why don't you erase the rootfs via u-boot cmdline? May 06 11:19:25 coredump: hmm, I did not realize that possibility. I added zeros, already :-) May 06 11:23:22 pavelm: some have accidentally flashed wrong u-boot image, file name is a bit confusing (v4b3 vs v3 or something like that) May 06 11:24:11 aloril: Yes... I've seen the "feel free to pick up your uboot" ridle. May 06 11:24:15 s/ridle/riddle/ May 06 11:24:16 pavelm meant: aloril: Yes... I've seen the "feel free to pick up your uboot" riddle. May 06 11:24:42 * CoreDump|home hugs the debug board May 06 11:26:07 next attempt, I added 3MB of zeros at the end. May 06 11:27:56 too bad, still the same oops. May 06 11:28:39 how much flash do we have? May 06 11:30:54 I can't find the u-boot command to erase the rootfs again, I'm afraid May 06 11:32:33 pavelm: GTA01Bv3# nand erase clean rootfs May 06 11:34:30 * pavelm does not remember how to get to uboot console :-( May 06 11:35:43 aha, minicom /dev/ttyACM0 May 06 11:35:59 press and hold 911-key while powering up neo should drop you into the menu. At this point u-boot is available via /dev/ttyACM0 May 06 11:36:28 (script) wiki RecentChanges: [[Wish_List_-_Hardware]] [[Flashing_openmoko]] [[Dfu-util]] [[Supported_microSD_cards]] May 06 11:37:21 yep, it looks like erase succeeded. May 06 11:40:00 bbl May 06 11:40:24 thanks for help, core. May 06 11:40:54 np May 06 11:45:10 good, erasing helped, now I am at 'calibrate touchscreen' phase. May 06 11:53:52 ...good, it boots. But I tried to enter terminal, and whole system crashed :-(. May 06 11:57:11 how does crashed openmoko look? May 06 11:57:58 balrog: just dead, does not react to anything. May 06 11:58:05 i've only seen lockups (which probably meant there was an Oops message in the console, but the console was not visible) May 06 11:58:08 what, no green screen of death?! May 06 11:58:08 ah okay May 06 11:58:22 pavelm: The very old u-boot has only 64MB usable. If you attempt to use more then you will overwrite the content of the lower RAM May 06 11:58:38 pavelm: Is it still the stock u-boot? May 06 11:58:44 and get an Oops May 06 11:59:00 alphaone: Yes, I still have stock uboot. But I get X locking up, not uboot. May 06 11:59:09 You should definitely upgrade then. May 06 11:59:18 pavelm: yes, that's most probably cause by u-boot May 06 11:59:34 *caused May 06 11:59:35 alphaone: Is uboot version really important once kernel is loaded? May 06 11:59:41 pavelm: Because u-boot is not using memory above the 64Meg limit. May 06 11:59:54 u-boot initialises the RAM banks, if it doesn't do that the kernel cannot use them either May 06 11:59:59 You could try booting with mem=64M option and see if that works May 06 12:00:15 Ok, that looks like an explanation. May 06 12:00:36 Is u-boot-gta01v3-r5 what I want on phase0 machine? May 06 12:00:58 pavelm: u-boot sets certain pins as address pins for the memory. Or rather it forgot to set the most significant pin in the old version. May 06 12:01:09 no. May 06 12:01:15 gta01bv3 May 06 12:01:59 i think you want the latest gta01bv3 build available May 06 12:02:34 yeah, basically May 06 12:03:42 Ok, u-boot-gta01bv3-r5_9877d7dcd1eebe61aa5d8b8ffe9c048ea426e6f6_0_1910.bin for me, then. May 06 12:04:51 And keep your finger on the aux button May 06 12:06:37 My openmoko survived uboot update; good. May 06 12:11:18 I still get no sound with the latest (may-01) software :-( May 06 12:11:23 I can do a call, but I can't hear anything. May 06 12:14:35 pavelm: are you running the magic alsa commands? May 06 12:15:04 robster: No, I'm not. I was hoping that would be fixed in may-01 version...? May 06 12:15:34 aha... so the software lets me call, even when the gsm is not registered to the network. May 06 12:15:46 Yes, theres "antenna ?" hint in the task bar. May 06 12:15:56 But no, it should not tell me that call is connected. May 06 12:16:19 i think the dialer doesn't work. May 06 12:16:56 the gsm integration is being rejigged. May 06 12:19:48 robster: Yes, looks so. I'll try to turn off pin at the test sim card. May 06 12:24:43 Will it be possible to run the openmoko without the binary gsm firmware? As I really have little use for a phone. May 06 12:27:12 I would actually be using it just as a PDA mostly and something to practicing developing on, just being a big Free software advocate I'd rather rip of the binary firmware even though it'd mean I lose some functionality. May 06 12:27:55 mukunda: The firmware actually runs on separate cpu. May 06 12:28:10 Mukunda: yes, OpenMoko is not being made as a phone specific distro May 06 12:28:15 mukunda: So just keep that cpu powered down. May 06 12:28:21 pavelm: ahh. May 06 12:29:02 So it's not actually user space firmware, like we've been seeing with a lot of new pc devices? May 06 12:29:31 Mukunda: if it runs on the host then it isn't firmware. May 06 12:29:45 i think there will even be PDAs using OpenMoko released later this year (supposedly), after Neo May 06 12:30:05 Ooh hm, that is what I'd actually be more interested in. May 06 12:30:49 Right now though, the openmoko seems to be cheaper and better than any PDA out there in that price range. May 06 12:33:23 If they can find a GPS chip that would allow a 100% Free software PDA then I'd buy two of the things just to support them. May 06 12:33:33 So hopefully they'll do it. May 06 12:34:51 mukunda: http://wiki.openmoko.org/wiki/Hammerhead_protocol -- feel free to help. May 06 12:37:20 uh, oh. Sound is not enabled in may-01 build May 06 12:39:20 Hm that's some impressive work on the hammerhead thing, well to me, but I'm pretty newbie. May 06 12:47:38 Mukunda: A problem is that the hammerhead chip is fundamentally better performance wise than most of the NMEA output chips. May 06 12:47:41 And cheaper. May 06 12:47:47 And lower power. May 06 12:47:56 So, with a GPL driver, it'd be ideal. May 06 12:48:03 Even without. It's not a kernel mode driver. May 06 12:48:33 It's a userspace thing that connects to the GPS serail port and GPSd May 06 12:48:44 You can chroot it in its own little box if you want. May 06 13:03:07 SpeedEvil: yeah, neat... If a GPL driver can be made that'd be awesome.. Shame the company isn't willing to help. May 06 13:11:13 Personally, I don't really care. As long as a working binary is shipped. May 06 13:11:42 If the binary driver can be chrooted, which it seems it can, and have assorted other walls put round it, there isn't a security worry. May 06 13:12:00 And if a GPL driver can be produced at leisure, that's fine IMO. May 06 13:13:54 I do care but yeah, it's a good tradeoff considering that it's very likely said GPL driver will be produced at leisure May 06 13:14:54 (meanwhile, I will be using the binary one only mildly grudgingly ;] ) May 06 13:19:21 I have problem with openocd and my broken Bv4 phone. Anyone that could help. May 06 13:20:29 holtmann: bv4... I do not even know such thing exists. May 06 13:22:30 holtmann: i'm not an openocd expert, but what's your problem? May 06 13:28:19 Seems there is a sea change regarding VoIP, what with the Verizon/Vonage verdict and AT&T/SMB turning-off VoIP customers... May 06 13:28:30 Make WiFi even more important... May 06 13:29:04 s/Make/Makes/ May 06 13:29:04 DukeOfURL_ meant: Makes WiFi even more important... May 06 13:32:11 In some places. May 06 13:33:14 mickeyl: I made the huge mistake to flash the v3 bootloader into v4 hardware. It is now a vibrating brick. May 06 13:33:24 uh oh May 06 13:33:25 We have a local cable provider (Comcast) that purposly introduces jitter into voice media streams May 06 13:33:49 holtmann: i recommend using the devirginator May 06 13:34:00 holtmann: this wraps openocd usage May 06 13:34:10 I am trying that. I have openocd connected, but deviriginator doesn't work. May 06 13:34:31 waiting for target halted... May 06 13:34:31 timed out while waiting for target halt May 06 13:34:31 rbp 0x33f80000 May 06 13:34:46 mww 0x32000040 0x32100000 May 06 13:34:46 error: target must be halted for memory accesses May 06 13:34:58 please check FPC seating May 06 13:35:11 pavelm: Yes it exists and I just bricked mine last week. May 06 13:35:21 mickeyl: Explain please? May 06 13:35:31 the FPC is very picky May 06 13:35:35 it must fully seat in the slots May 06 13:35:42 check whether it sits correctly May 06 13:36:28 (script) wiki RecentChanges: [[Wish_List_-_Hardware]] [[Flashing_openmoko]] May 06 13:36:46 * philippe is glad to see he is not the only one to be stupid enough to flash wrong bootloaders :-) May 06 13:37:15 I was pure accident since I was previously working on the v3 hardware. May 06 13:37:45 holtmann: trust me, I know how it feels :-) May 06 13:38:23 Crap. No I openocd doesn't start anymore. May 06 13:40:47 Bv4 is phase0 + fix the power consumption during sleep + fix the GSM sidetone? May 06 13:41:22 Error: jtag.c:1236 jtag_examine_chain(): JTAG communication failure, check connection, JTAG interface, target power etc. May 06 13:41:30 holtmann: this is almost always a power problem May 06 13:41:35 I get this now all the time. Previously it was working. May 06 13:41:36 holtmann: to properly reset May 06 13:41:43 a) unconnect everything May 06 13:41:46 b) remove battery May 06 13:41:50 Did that two times now. May 06 13:41:52 c) connect FPC May 06 13:41:59 d) press power button for 2 seconds May 06 13:42:14 e) run openocd May 06 13:42:28 c.2) reattach debug board USB May 06 13:42:48 oh, and be sure to use a powered battery.... May 06 13:43:04 openocd now works again. May 06 13:43:22 good May 06 13:43:27 try reset halt May 06 13:43:31 before running devirg May 06 13:43:42 if it doesn't halt then you still have connection problems May 06 13:44:15 target not halted May 06 13:44:15 wait_halt May 06 13:44:15 waiting for target halted... May 06 13:44:15 timed out while waiting for target halt May 06 13:44:15 This is the problem. May 06 13:44:28 hmm May 06 13:44:42 i'm lost there May 06 13:45:11 LaF0rge and wpwrak are the openocd experts May 06 13:45:17 might ping them May 06 13:45:19 I used the openocd from SVN. Should be okay since all of Werner's patches seem to be integrated. May 06 13:45:19 Warning: arm7_9_common.c:1378 arm7_9_resume(): target not halted May 06 13:45:19 Error: target.c:1334 handle_wait_halt_command(): timed out while waiting for target halt May 06 13:45:20 Error: breakpoints.c:134 breakpoint_remove(): no breakpoint at address 0x33f80000 found May 06 13:45:22 From openocd. May 06 13:46:46 No difference. May 06 13:46:46 Actually it does reset halt as first command. So that is a no op. May 06 13:47:12 darn May 06 13:47:30 got 1.9 of that lib? May 06 13:47:59 Yes. May 06 13:48:07 The halt is the problem. May 06 13:49:36 let me upload my configuration to pastebin to compare May 06 13:49:45 that'll take some minutes (booting linux machine) May 06 13:50:52 Do you think that I completely bricked it or is it possible to revive it. May 06 13:51:02 it is always possible to revive it May 06 13:51:13 Btw. the vibrating brick was not a joke. Once it is on, it vibrates. May 06 13:51:38 might disconnect the vibrator for now May 06 13:51:53 problem are GPIO changes between v3 and v4 May 06 13:52:48 Do I need a different version of openocd (other than latest SVN) or a different debug board. May 06 13:52:50 pastebin.ca/473874 May 06 13:54:27 working_area is commented out, but otherwise it is the same. May 06 13:55:17 hmm. perhaps the FPC is broken. please mail to -devel. i'm really out of my depth w/ openocd May 06 13:55:33 I'm hoping wpwrak or LaF0rge come up later today May 06 13:56:03 it's 22 in taipeh and 11 in buenos aires May 06 13:56:08 should be a good time May 06 13:57:40 Let me try another cable. I have three of them with me. However the debug board still has the old VID/PID. May 06 13:57:50 But it is the Debug Board 2. May 06 13:58:18 might flash the eeprom May 06 13:58:36 i wouldn't trust Taiwan to have it prepared correctly... May 06 13:58:38 *cough* May 06 13:59:51 No difference. Okay where is the howto to flash the debug board. May 06 14:00:24 http://wiki.openmoko.org/wiki/Debug_Board May 06 14:02:56 mickeyl: have you got a few minutes to give me access to the artwork repository? May 06 14:03:49 wansti: run htpasswd -bm May 06 14:03:54 and mail the output to me May 06 14:03:59 okay May 06 14:04:16 (mickey@openmoko.org) May 06 14:07:08 mickeyl: Even with the flashed debug board. Same issue. May 06 14:07:22 holtmann: ok, sorry, then you need to wait for werner or harald May 06 14:09:07 T: Bus=02 Lev=02 Prnt=116 Port=00 Cnt=01 Dev#=117 Spd=12 MxCh= 0 May 06 14:09:07 D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 8 #Cfgs= 1 May 06 14:09:07 P: Vendor=1457 ProdID=5118 Rev= 4.00 May 06 14:09:07 S: Manufacturer=?????????????? May 06 14:09:07 S: Product=??????????????????????? May 06 14:09:10 Ups. May 06 14:09:41 mickeyl: done May 06 14:12:45 hey holtmann May 06 14:12:57 Hi. May 06 14:13:16 wansti: key added. recheckout the artwork directory via https and you should be able to rw May 06 14:14:29 wansti: i didn't forget to record Sabine btw., but since we're right within moving, we didn't have time. I will send you the recording in early June May 06 14:14:49 koen: My v4 hardware is still broken :( May 06 14:14:55 mickeyl: okay, nice :-) May 06 14:15:14 holtmann: I'm still on v3 May 06 14:15:21 holtmann: attending GUADEC? May 06 14:16:12 I should, but I would need a sponsor for hotel and flights. So maybe not. May 06 14:17:35 I'm in the UK anyway, so my flight to birmingham is £0.04 May 06 14:17:43 (+£20 tax) May 06 14:19:43 i'll commit my stuff under /trunk/src/target/OM-2007/artwork/sounds... that okay? May 06 14:19:52 yep May 06 14:21:31 holtmann: any sign of a 3.10 release yet? May 06 14:22:07 We have to replace one fork() with a queue integrated into the main loop. Other than that, it's ready. May 06 14:22:40 cool May 06 14:23:27 openmoko: 03wansti * r1916 10/trunk/src/target/OM-2007/artwork/sounds/ringtone_thereisnophone.mp3: May 06 14:23:27 openmoko: first ringtone (will be replaced with a better recording soon, this is May 06 14:23:27 openmoko: mainly for testing svn access :-)) May 06 14:23:37 tadaaa May 06 14:23:38 :) May 06 14:23:41 :-) May 06 14:24:27 bye, everyone. May 06 14:25:01 cu pavelm May 06 14:30:44 LaF0rge: I can halt the chip using openocd. May 06 14:30:49 s/can/can't/ May 06 14:30:49 holtmann meant: LaF0rge: I can't halt the chip using openocd. May 06 14:31:12 And I just put the phone back together :( May 06 14:32:03 LaF0rge: I made the mistake and flashed the v4 boot loader into the v3 hardware. May 06 14:33:09 holtmann: you have to do a bit of trickery May 06 14:33:23 Can you guide me through it. May 06 14:33:41 holtmann: e.g. push down reset button on debug board, then power up the device, then issue 'halt' command on openocd, then release reset button May 06 14:34:30 Just a few seconds to setup it up again. I just packed everything back together. May 06 14:35:16 if it doesn't work, just keep retrying ;) May 06 14:35:47 and as new bootloader, make sure you use bv3 (the B is important May 06 14:36:58 What is the best download place for all the latest stuff. May 06 14:37:28 holtmann: well, since you have a broken device now, the good old lowlevel_foo + breakpoint method is the only way May 06 14:37:54 What link should I use? May 06 14:38:08 link for description or link for u-boot? May 06 14:38:18 http://buildhost.openmoko.org/tmp/deploy/images/u-boot-gta01bv4-r5_9877d7dcd1eebe61aa5d8b8ffe9c048ea426e6f6_0_1910.bin May 06 14:39:10 oh, sorry May 06 14:39:12 that's Bv4 May 06 14:39:23 http://buildhost.openmoko.org/tmp/deploy/images/u-boot-gta01bv3-r5_9877d7dcd1eebe61aa5d8b8ffe9c048ea426e6f6_0_1910.bin May 06 14:39:34 No. It is v4 hardware. May 06 14:40:26 22:32 LaF0rge: I made the mistake and flashed the v4 boot loader into May 06 14:40:26 the v3 hardware. May 06 14:40:38 I meant the other way around. Sorry. May 06 14:40:43 ah, ok May 06 14:44:33 let me know once you've managed to halt the cpu May 06 14:44:43 It is halted now. May 06 14:44:48 Holding reset did the trick. May 06 14:45:11 Now ./devirginate should be enough. May 06 14:46:06 yes, if you give it correct filenames for uboot/kernel/rootfs, it should work May 06 14:49:42 Stage 1 complete. No stage 2. May 06 14:49:44 Error: arm7_9_common.c:562 arm7_9_execute_sys_speed(): timeout waiting for SYSCOMP & DBGACK, last DBG_STATUS: 0 May 06 14:49:54 I get this. May 06 14:51:20 openmoko: 03wansti * r1917 10/trunk/src/target/OM-2007/artwork/sounds/ (6 files): added Rosegarden source files for all (upcoming) sounds May 06 14:52:29 mh, strange. that's one of those jtag issues. should be gone with recent openocd/libusb May 06 14:52:36 in any case, after stage 1 you have u-boot there May 06 14:52:41 openocd is from CVS. May 06 14:52:52 i.e. you should be able to use dfu-util for kernel/rootfs May 06 14:53:35 How do I make it boot now. May 06 14:54:41 It doesn't show up on the USB bus. May 06 14:54:50 does the screen light up/ May 06 14:54:50 ? May 06 14:54:55 No May 06 14:55:01 i.e. press AUX button plus power for 5 sec/ May 06 14:55:01 ? May 06 14:55:14 Error: arm_jtag.c:38 arm_jtag_set_instr_error_handler(): setting the new JTAG instruction failed, debugging is likely to be broken May 06 14:55:45 hm, never seen that jtag error. what do you see on the serial port? May 06 14:55:57 Which serial port? May 06 14:56:21 holtmann: the debug board exports two serial devices: the first one is JTAG used by openocd, the seond one is the serial console of the neo May 06 14:56:47 holtmann: you need to load ftdi_sio with appropriate vendor/product parameters to make that work, as described per wiki May 06 14:57:10 Nothing on it. May 06 14:57:47 mh. strange. and you are sure you have a Bv4 hardware? May 06 14:58:08 Bv4 should be the non-bluetooth-device that FIC sent you very recently in accident May 06 14:58:19 No. I have a new one. May 06 14:58:25 ah, ok. May 06 14:58:29 Wait a second. I might got it. May 06 14:59:05 Dfu download is now in progress. No idea what I changed. May 06 14:59:11 mh. May 06 14:59:20 Only unplugged the debug board. May 06 14:59:40 ah. maybe openocd got somehow stuck and kept the board in halt via jtag May 06 14:59:43 Hopefully it also flashed the boot loader now. May 06 14:59:55 openocd with libusb < 0.9 is quite unreliable May 06 15:00:18 And btw. they forgot to send me another battery. So I have now two devices, but only one battery. May 06 15:00:54 mh, too bad. I'm not sure if there are still some in berlin. will ask my office assistant to check and mail you one May 06 15:01:11 I am in Sevilla anyway. So won't help me just right now. May 06 15:01:43 when are you back, and how long? May 06 15:01:52 I will be back next Saturday. May 06 15:06:48 I have back the boot loader. At least I think so. Where can I find the latest snapshot to flash into it. May 06 15:06:52 LaF0rge: there's still something wonky in the 2.6.21 kernel, I can get gsmd to register on the network nor control the backlight May 06 15:07:12 http://buildhost.openmoko.org/tmp/deploy/images/ ? May 06 15:07:42 holtmann: There is not snapshot per se, you can just manually select the lastest uboot/uImage/openmoko-devel-image May 06 15:08:26 holtmann: http://buildhost.openmoko.org/tmp/deploy/images/?C=M;O=D helps May 06 15:08:52 koen: as long as the backlight driver is not compiled in you won't be able to control it :) May 06 15:10:48 Do you also have link for the splash. It seems I lost it somehow. May 06 15:11:09 holtmann: i think you need to use devirginator to do the conversion May 06 15:11:27 Does dfu upload work. Then I can use the one from my v3. May 06 15:11:35 holtmann: devirginator downloads the original png and converts it into the correct format, and then saves it to tmp/ May 06 15:12:04 dfu upload works depending on your u-boot version. before April 19 I think it was working, and after MAy 03 it is working again May 06 15:12:29 Okay. I would boot without, right? May 06 15:12:38 What about the config area. May 06 15:13:01 yes, it boots without the splash May 06 15:13:38 the environment? I think without environment it would be missing some boot parameters. Don't know if the compiled-in default is sufficient May 06 15:14:30 also, if you manually install everything, you might need to set the dynenv partition marker. (if 'dynenv get' returns 'no marker' then use 'dynenv set u-boot_env' May 06 15:14:48 What is dynenv? May 06 15:15:17 holtmann: the dynamic (per-device specific) start sector of the envirohment partition, which is stored in the OOB area of the first NAND flash sector May 06 15:15:36 if you erase/re-write u-boot manually, then that marker will be erased, too May 06 15:15:59 I flash the rootfs, kernel and uboot now. May 06 15:16:00 if you use DFU to flash a new u-boot, it saves that marker and re-writes it after finishing the write to the u-boto partition May 06 15:16:55 rtp: CONFIG_BACKLIGHT_LCD_SUPPORT=y May 06 15:17:57 Okay. My boot loader is from May 5th. So that seems to be okay, but it doesn't boot and the menu only shows boot and factory reset. May 06 15:18:19 then the environment is missing May 06 15:18:22 or the environment marker May 06 15:18:36 I flashed the u-boot via DFU. May 06 15:18:45 check it as described above by issuing 'dynenv get' May 06 15:18:46 koen: you need BACKLIGHT_GTA01 which depends on BACKLIGHT_DEVICE and BACKLIGHT_DEVICE has been removed May 06 15:19:18 rtp: there's no such config option May 06 15:19:19 holtmann: yes, but before using dfu, you [rather: devirginator] flashed it once without dfu, i guess (jtag) May 06 15:19:37 So what should I do? May 06 15:19:56 rtp: http://rafb.net/p/3Wc10r59.html May 06 15:20:23 holtmann: try issuing 'dynenv get' on the u-boot command line May 06 15:21:29 koen: It's not there because the Kconfig entry depends on the now dead BACKLIGHT_DEVICE entry May 06 15:21:43 holtmann@aeonflux:~$ cu -l /dev/ttyACM0 May 06 15:21:43 cu: open (/dev/ttyACM0): Input/output error May 06 15:21:43 cu: /dev/ttyACM0: Line in use May 06 15:21:46 Argh. What the f* May 06 15:21:58 holtmann: chown uucp.uucp /dev/ttyACM0 May 06 15:22:16 holtmann: cu is very baroque and has strange ideas on device ownership May 06 15:22:34 Same problem. May 06 15:23:17 but the device actually exists? May 06 15:23:27 try minicom May 06 15:23:35 or picocom May 06 15:23:40 I wonder if all unit specific data are at the dynenv part..like serial numbers of chips inside May 06 15:24:24 Any special extra parameter? May 06 15:24:35 when i was playing with Palm i found serial number in the middle of flash May 06 15:24:37 holtmann: no, not required. baudrate doesn't matter May 06 15:24:38 It somehow connects, but I can type on it. May 06 15:26:53 then the console might be switched to serial May 06 15:27:03 That is funny. The factory reset option actually boots the kernel. May 06 15:27:07 try accessing the USB1 device (or whatever the 2nd serial of the debug board might be) May 06 15:27:12 :) May 06 15:27:17 yes, it's supposed to do May 06 15:27:34 Okay, but boot powers the device off. May 06 15:27:53 So if I have it booted, does that help me? May 06 15:29:25 well, you then have the device running. May 06 15:29:35 guess that's what you wanted, after all. May 06 15:29:46 in any case, you should still fix your environment May 06 15:29:54 i.e. get access to the u-boot console at some poitn May 06 15:30:56 let me try that. So I connect the debug board and get into the boot loader. Then one of ttyUSB should give the serial console of the boot loader. May 06 15:31:29 yes May 06 15:32:56 I only get terminal ready, but can't use it. May 06 15:33:12 Daymn May 06 15:33:17 is it for sale yet??? May 06 15:33:18 humok: 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 06 15:34:02 holtmann: mh, it should definitely respond to commands on either the serial or the usb. there is no other setting May 06 15:34:15 10th may May 06 15:35:30 humok/aloril: Bv4 production is getting transferred into mainland china factory right now May 06 15:35:45 I see garbage on ttyUSB once I press reset on the debug board. May 06 15:35:50 LaF0rge: thats good news i take it? May 06 15:35:54 ttyUSB1. May 06 15:35:58 so it should be availiable soon May 06 15:36:00 right May 06 15:36:02 holtmann: oh, on ttyUSB1 you need 115200 bps May 06 15:36:08 ive been dieing for this fone May 06 15:36:19 seems untouchable by others May 06 15:36:26 holtmann: there the baudrate matters. only on usbtty it donsn't. May 06 15:36:48 Ah. Here we go. May 06 15:36:55 holtmann: 'dynenv get' May 06 15:37:04 GTA01Bv4 # dynenv get May 06 15:37:04 0x00040000 May 06 15:37:28 (script) wiki RecentChanges: [[Wish_List_-_Hardware]] [[Wishlist:Unlikely]] May 06 15:39:31 holtmann: oh, so the marker is there May 06 15:39:37 only your environment seems erased :( May 06 15:39:51 LaF0rge: the production transfere means that it is going to mass production? May 06 15:39:53 i.e. the actual content of the environment partition. May 06 15:40:07 How do I get it back? May 06 15:40:21 magon: well, technically MP but not in regular MP quantities (we are far from being mass consumer compatible) May 06 15:40:22 And how do I switch it to usbtty. May 06 15:40:48 LaF0rge: ok but harwdare issues are solved? May 06 15:40:54 holtmann: setenv stderr usbtty; setenv stdout usbtty; setenv stdin usbtty May 06 15:41:23 holtmann: then 'saveenv' (from usbtty since the changes should be effective immediately) May 06 15:41:33 LaF0rge: great news! will there be pre-orders? May 06 15:41:45 aloril: not sure, I'm R&D, not sales ;) May 06 15:41:47 No. Still using the USB1. May 06 15:42:08 holtmann: yes, this is because there were no such variables in the environment before. May 06 15:42:22 holtmann: usually it becomes in effect immediately, but not if your env was corrupted ;) May 06 15:43:44 LaF0rge: plan is to product Bv4 in factory than make HW refresh and move it to factory again? May 06 15:43:44 It does. For a second I had stdin on USB1 and stdout on ACM0 May 06 15:44:05 So what else do I need to get this back into a sane state. May 06 15:44:24 holtmann: it's not easy manually May 06 15:44:35 holtmann: usually devirginator takes care of this May 06 15:45:13 GTA01Bv4 # printenv May 06 15:45:13 bootargs= May 06 15:45:13 baudrate=115200 May 06 15:45:13 bootargs_base=rootfstype=jffs2 root=/dev/mtdblock4 console=ttySAC0,115200 console=tty0 loglevel=8 May 06 15:45:13 mtdids=nand0=neo1973-nand May 06 15:45:14 mtdparts=mtdparts=neo1973-nand:0x00040000(u-boot),0x00004000(u-boot_env),0x00200000(kernel),0x000a0000(splash),0x03d1c000(rootfs) May 06 15:45:17 bootdelay=604800 May 06 15:45:19 splashimage=nand read.e 0x32000000 splash 0x1000; unzip 0x32000000 0x33d00000 0x96000 May 06 15:45:21 usbtty=cdc_acm May 06 15:45:23 partition=nand0,0 May 06 15:45:25 mtddevnum=0 May 06 15:45:27 mtddevname=u-boot May 06 15:45:29 pcf50606_int1=0x4a May 06 15:45:31 stderr=usbtty May 06 15:45:33 stdin=usbtty May 06 15:45:35 stdout=usbtty May 06 15:45:37 Environment size: 506/16380 bytes May 06 15:45:41 that is how it looks currently. May 06 15:45:43 ~lart people_who_dont_use_pastebin May 06 15:45:43 * apt steals people_who_dont_use_pastebin's mojo May 06 15:47:05 LaF0rge: Having it boot would be enough right now. May 06 15:47:19 holtmann: well, 'setenv bootdelay 10' should do half of the trick May 06 15:47:42 Why? May 06 15:48:10 oh and 'setenv bootcmd setenv bootargs \${bootargs_base} \${mtdparts}\; nand read.e 0x32000000 kernel\; bootm 0x32000000 May 06 15:48:12 ' May 06 15:48:40 holtmann: because bootdelay of 604000 seconds is not good for auto-boot :) May 06 15:49:01 *nod* May 06 15:49:10 Oh, I see. The value was why to high. May 06 15:49:15 No it is not. May 06 15:49:31 It still doesn't boot. May 06 15:49:44 And the boot option in the boot loader turns it off. May 06 15:50:26 * aloril2 just hopes that he manages to order Neo1973 enough fast May 06 15:52:28 there isn't any competition ;-) May 06 15:55:11 holtmann: basically, if 'boot' fails to boot it, it turns off May 06 15:55:28 holtmann: boot it into the menu (Aux+power), then access usbtty and issue 'boot' and see what happens May 06 15:56:05 also, use printenv to verify bootcmd and bootdelay variables May 06 16:00:37 Okay. Seems to work now. Thanks a lot. May 06 16:01:20 what exactly meand reset to Factory default? May 06 16:02:05 magon: not yet implemented. it basically issues a hardcoded boot command and boots into the kernel with a particular option, where some not-yet-implemented userspace program shall take care of resetting everything to defaults May 06 16:02:37 ok but you have to not reflash it May 06 16:03:45 what about implementing setting reasonable dynenv as factory default? May 06 16:04:44 LaF0rge: Btw. the power_on for the Bluetooth dongle doesn't work anymore. May 06 16:07:47 Okay. the Boot menu now works, but autoboot not. What is wrong there? May 06 16:13:19 autoboot is only active if you don't press the aux key May 06 16:13:23 you know that, i guess ;) May 06 16:13:40 holtmann: apart from that, the only thing i can think of is bootdelay May 06 16:15:56 I think my config storage is completely messed up. May 06 16:17:06 holtmann: what you sent above actually semed quite reasonable May 06 16:17:31 Yeah, but now I get some garbage when I call printenv. May 06 16:17:49 bad blocks? May 06 16:18:41 bootdelay=10 May 06 16:18:41 6000 May 06 16:18:41 stderr=usbtty May 06 16:18:41 stdin=usbttootargs ${bootargs_base} ${mtdparts}; nand read.e 0x32000000 kernel; bootm 0x32000000 May 06 16:18:41 partition=nand0,0 May 06 16:18:41 mtddevnum=0 May 06 16:18:43 mtddevname=u-boot May 06 16:18:45 Stuff like this. May 06 16:19:05 See line with stdin May 06 16:20:49 Does stage 2 of devirginate also fixes the config or do I need stage 1 for that? May 06 16:25:23 holtmann: the garbage is probably usb related May 06 16:25:39 holtmann: usbtty has some ringbuffer corruption resulting in bogus printouts if you display a lot of text May 06 16:25:47 holtmann: sorry for that, nobody ever got around fixing that May 06 16:26:11 commands that return less text, such as 'printenv stdin' should give more reliable output May 06 16:27:34 holtmann: please talk to wpwrak about any further issues, it's getting late here so I'll go to bed soon, sorry May 06 16:27:48 werner (wpwrak) is online right now, just talking to him May 06 16:29:34 I am off for a guided tour and dinner. May 06 16:29:56 I can at least boot it now. However the power driver for Bluetooth is broken. May 06 17:05:25 has anybody tested 4GB cards yet? May 06 17:05:45 Are there any? May 06 17:06:56 microSD I mean. May 06 17:40:06 I read somewhere that the >2GB SD cards did not have larger capacity but were using compression. May 06 17:47:10 +mhentges@oe-neo1973:..73/out/tmp.dev_om/deploy >du images/ May 06 17:47:10 926M images/ May 06 17:47:37 I guess the OM deploy dir needs a "little bit" post-processing.... May 06 17:50:02 SpeedEvil: somebody refered to high 4GB price compared to 2GB, but maybe it was plain SD and not microSD May 06 17:50:27 I suspect so. May 06 17:50:33 I've not seen 4G microSD May 06 17:50:37 2G is rare. May 06 17:51:25 http://www.instructables.com/id/EM1O9XJF0YPF7M5/ Too cool. May 06 17:51:31 The perfect neo accessory. May 06 17:51:36 Or maybe not. May 06 17:52:08 there's a limit of SD sizes imposed by the protocol, from my reading of the specs that was 4 GB but i've heard 2 GB from people May 06 17:52:22 IIRC. May 06 17:52:30 1G is the intended limit. May 06 17:52:44 2G can be gotten either by bending the protocol a little, or going SDHC May 06 17:52:49 >2G has to be SDHC May 06 17:53:30 from my reading 2G wouldn't require any bending May 06 17:53:44 1G is the max with 512 byte blocks but blocks can be up to 2 K May 06 17:54:36 SpeedEvil: 2g microsd is widely available and cheap May 06 17:54:53 Slightly rarer I mean - last time I looked. May 06 17:54:56 :-O that mouse is disturbing May 06 17:54:56 4g microsdhc is expensive (twice the price per GB) and not really available May 06 17:54:58 thogugh it was a while. May 06 17:55:20 is microSD 32mm x 24mm? May 06 17:55:26 No. May 06 17:55:29 more like 20x10mm May 06 17:55:36 About 11*18*1 or so May 06 17:56:08 What is this one I'm holding? 32x24mm May 06 17:56:17 SD? May 06 17:56:18 I think that's probably a SD card May 06 17:56:48 SD or MMC, they look the same :) May 06 17:56:57 Most SD cards are physically thicker than MMC cards. SD cards generally measure 32 mm × 24 mm × 2.1 mm, May 06 17:57:04 http://en.wikipedia.org/wiki/Secure_Digital_card May 06 17:57:13 MMC = evil May 06 17:57:18 the MMC i had was also 2.1 mm thick May 06 17:57:26 mmc is less evil than SD May 06 17:57:51 Yeah. May 06 17:58:29 still very evil, though May 06 17:58:34 true enough May 06 17:58:50 SD protocol is less kludgy May 06 17:58:56 Compactflash isn't very evil. May 06 17:59:00 Too large though. May 06 17:59:29 I can't understand introducing SDHC. May 06 17:59:34 With a maximum size of 32G. May 06 17:59:37 CF is kinda braindead May 06 17:59:51 using full ATA spec for a flash card is way over the top May 06 17:59:52 raindead != evil though. May 06 18:00:13 USB card would be good May 06 18:00:16 640kB is enough for everyone! May 06 18:00:23 damn right it is May 06 18:00:43 * CoreDump|home is not everyone May 06 18:00:57 buz: thanks to using full ata it's 100% compatible with microdrives May 06 18:01:13 yeah but the connector is rather unwieldy May 06 18:01:27 it definitely doesnt suit small devices May 06 18:01:47 palm lifedrive has one, it works ok May 06 18:02:09 although most users i know (two) replaced the cards with microdrives already May 06 18:02:20 CF was _the_ format back in the days May 06 18:02:26 yeah, back many days May 06 18:02:33 my digicam still uses them May 06 18:02:36 it is still *the* format in pro photography May 06 18:02:38 so does mine May 06 18:02:44 spec says there's DRM on the SD card. I wonder how that works? May 06 18:02:51 pro photography doesnt use small equipment either May 06 18:03:03 i wonder why people buy microdrives these days May 06 18:03:06 flash is cheap May 06 18:03:12 and less hassle May 06 18:03:19 supposedly MD is faster May 06 18:03:34 could we use DRM to protect PIM data? May 06 18:03:37 md doesn't wear out due to excessive writes May 06 18:03:50 CF has write levelling iirc May 06 18:03:57 but since CF is essentially free nowadays May 06 18:04:08 koen: I heard MD's die _way_ faster than conventional HDDs so that point would be moot May 06 18:04:21 wearing out a 4GB CF card is kinda hard May 06 18:04:39 buz: flash cells are getting less writecycles these days May 06 18:04:46 maybe if you keep doing dd if=/dev/urandom of=CFcard May 06 18:04:54 you used to get 100k cycles before breakdown, with MLC only 10k May 06 18:05:04 i heard the opposite claim too May 06 18:05:32 look up the specs for the flash technology used May 06 18:05:54 read beyond the "OMG SIZE!!!" propaganda and you see how bad the write-cycle count is with MLC May 06 18:06:11 if the write leveling algorithm is any good, i can live with 10K writes May 06 18:06:35 i rather get cheap but big flash May 06 18:06:47 its not like i wont replace the card in due time May 06 18:07:27 * koen is happy with the 4gb flash card + 4gb microdrive backup for his cameras May 06 18:07:58 * koen firewire CF reader May 06 18:36:41 hello. May 06 18:49:27 Hi! May 06 18:50:03 ...so I've realized my sirfII gps receiver is crap (does not even respond to sirf commands properly) and am now reading opensource GPS documentation. May 06 18:50:08 ...quite a lot of reading ;_) May 06 18:55:48 the only thing in the topic is where or how to get it ;-) May 06 18:56:03 s/missing in the topc/ May 06 18:58:34 thats because its not known yet May 06 19:01:05 so rather late may if it still has to be decided/organised? May 06 19:01:06 kwadronaut: 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 06 19:07:20 anyone there? May 06 19:07:20 Clay: 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 06 19:08:33 yep. May 06 19:10:22 so, i want to start progamming, because i want a neo1973, but i dont know where to start May 06 19:11:41 can anyone help me start programming May 06 19:12:10 Clay: OpenMoko is basically normal Linux distribution May 06 19:12:42 i know, but im really new, i dont mean to be a menice May 06 19:13:41 do you now C and/or Python? May 06 19:14:03 i know a little of both, more python than C May 06 19:15:27 what about gtk? May 06 19:15:56 i cant get my compiler to compile gtk May 06 19:17:42 usually gtk is included in Linux distribution, do you mean you can't compile gtk using program? (possible solution: install gtk-devel) May 06 19:18:00 will OpenMoko be carrier agnostic? May 06 19:18:39 hadak: Yep. May 06 19:18:51 HaDAk: should be May 06 19:18:58 unfortunately, i have windows, i have a ubuntu and slax live cd tho, will that work? May 06 19:19:16 thank God, i can finally get away from the mainstream phones. so limiting! May 06 19:19:42 hadak: Mainstream phones are carrier agnostic, too :-). That's called unlocking. May 06 19:19:46 only when neo is released May 06 19:19:47 i read that availablility will start in june/july...and mainstream production starts in aug. is this still the projected timeline? May 06 19:20:09 (though FIC probably is working to also get distribution deals with carriers for mass market release) May 06 19:20:17 pavelm: i understand that. it is legal, however phone companies have contingencies that allow them to deny you service if you're using an unlocked phone. May 06 19:20:41 my best guess for P1 is this: May 06 19:20:47 counter May 06 19:20:47 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, a day 04:39:10 (8.194 +-3.5 days) (1099;195) May 06 19:21:02 hadak: Just buy a phone from someone else than carrier. And in my country, carriers themselves sell unlocked phones. May 06 19:22:19 sounds good. this will be available online only? May 06 19:24:21 aloril2: will a linux live cd with slax or ubuntu work for compiling gtk? May 06 19:24:32 Clay: it's easier to install Linux and develop there than to develop under windows (which is possible too) May 06 19:25:07 Clay: look for colinux or something like that (Linux under Windows) May 06 19:25:23 allright, thanx May 06 19:26:33 Clay: Maybe using some kind of virualized install of Linux would be simpler. Check http://virtualbox.org May 06 19:28:17 cygwin is yet another option. but once you start compiling, better be running 'just' *nix May 06 19:28:19 allright, im going, but i have dial-up, so it's hella slow May 06 19:28:35 especially if you have an older box May 06 19:29:26 counter May 06 19:29:26 http://wiki.openmoko.org/wiki/Neo1973 : >= 2007-05-10 (uncertainty is a guess) P1: a week, a day 04:30:32 (8.188 +-3.5 days) (1100;195) May 06 19:37:19 so, im trying virtualbox, would i need a distro of linux? May 06 19:37:28 (script) wiki RecentChanges: [[Hammerhead_protocol]] May 06 19:41:38 Clay: Yes. It's just like you would install it on your computer but instead you install it on the virtual machine. You workstation should have enough horsepower thought otherwise it will be awfully slow. May 06 19:41:47 Clay, you need some OS to run on a virtual machine May 06 19:42:30 allright, any suggestions which distro? May 06 19:42:58 actually i don't think the performance hit is that great. maybe some for disk access though since this filesystem is emulate May 06 19:43:14 d May 06 19:43:53 peepsalot: Mostly depends on the amount of memory. May 06 19:45:18 Clay: Debian or Ubuntu might be a sound choice but that's really a matter of taste I guess. May 06 19:46:15 i havent tried debian, but i have a ubuntu live cd, i can use that to install it on my virtualbox, right? May 06 19:46:58 provided that it's an install CD May 06 19:47:05 Clay: yes, then you should go with Ubuntu. May 06 19:47:18 but you don't have to install ubuntu, you can use the live-cd only May 06 19:47:20 ok, thanx again everybody May 06 19:48:14 i'd use qemu anyway May 06 19:48:30 qemu? May 06 19:48:57 yeah, virtual box is a fork of qemu May 06 19:49:20 oh okay, im really new, sorry May 06 19:49:35 that's no problem, virtual box will work too May 06 19:52:57 balrog-kun: AFAIK Virtualbox uses qemu for certain things but is not a fork. May 06 19:53:58 i have a question, i have an external 60-gig hdd... May 06 19:54:25 balrog-kun: see http://www.virtualbox.org/wiki/Developer_FAQ May 06 19:54:29 Kensan: over 60% of the code in virtual box is our code May 06 19:54:45 could i install linux on that and have my bios boot that directly? May 06 20:01:37 Clay, yes if your BIOS supports booting from USB - assuming your external HD is USB. May 06 20:02:16 yes it is, and my bios supports it, would that be better than virtual box? May 06 20:02:41 In some ways, yes. May 06 20:03:03 probably faster and more reliable May 06 20:03:07 It's a 'real' running system, and you don't need to configure stuff in complex ways. May 06 20:03:20 OTOH - you can't run normally at the same time. May 06 20:03:32 If you are "serious" about development I think you should have a "native"/full Linux installation at some point. May 06 20:04:24 True. May 06 20:04:46 in the end OpenMoko is itself a Linux distro... May 06 20:04:51 but that's just my 2 cents. May 06 20:05:08 Not used. May 06 20:05:20 oops May 06 20:05:22 ok, so is programming for OpenMoko hard? May 06 20:05:42 Not especially harder than any other linux programming. May 06 20:06:03 It depends mostly on what you generally know about programming, linux, etc May 06 20:06:21 well, i havent done much other linux programming, but i want to learn May 06 20:07:04 Other than libs, most of it should transfer over. May 06 20:07:25 If you've been doing windows programming. May 06 20:07:53 It's more about finding your way around, e.g. how to set up the development environment etc. but if you are willing to learn then this is a great chance May 06 20:08:05 Yeah. May 06 20:08:10 not a whole lot of windows programming either besides some tutorials online May 06 20:11:16 You've probably got a long way to go before you can create reasonable apps then, but I suspect you knew that. May 06 20:11:41 yeah May 06 20:12:03 besides, you do not need to create reasonable apps ;) just do what you want to do May 06 20:12:22 I mean more than 'helo wrlf'. May 06 20:12:46 Useful apps can of course be very very small. May 06 20:14:07 definately May 06 20:15:04 Putting Ubuntu on your external HD and setting up the development env would be a good start I guess. May 06 20:15:42 what do u mean by developement env? the libraries and such? May 06 20:16:21 ~lart the still non-functional GPS May 06 20:16:22 * apt cuts the still non-functional GPS into thin stripes May 06 20:16:37 using free software on the PC would be a good start to freeing your phone etc May 06 20:16:59 So has anyone here actually got an openmoko yet? May 06 20:17:10 a few of the devs yes May 06 20:17:17 my phone is basically my last non-OSS device May 06 20:17:19 as for the unwashed masses, no May 06 20:17:32 Ah.. Are there any decent photos of it out there yet? May 06 20:17:44 sure, try the website May 06 20:17:59 or flickr May 06 20:18:04 wiki, planet..? May 06 20:18:07 some 4megapixel shots on flickr May 06 20:18:07 wiki.openmoko.org should point to photos May 06 20:18:20 Danke. May 06 20:18:20 I do have phase0 device here. May 06 20:18:21 planet is rather barren ;) May 06 20:18:22 Clay: Set up the libraries etc so you can build OpenMoko, there's lots of information in the wiki May 06 20:19:21 !praise MokoMakefile May 06 20:19:22 * cdbot2 bows down to MokoMakefile and chants, "I'M NOT WORTHY!!" May 06 20:23:12 4mpixel? May 06 20:23:17 8 mpixel! May 06 20:23:18 alright, so i use the instructions of MokoMakeFile? May 06 20:23:24 hehe May 06 20:24:05 Clay, yes, once you have your linux running follow instructions for MokoMakeFile (http://wiki.openmoko.org/wiki/MokoMakefile) May 06 20:24:22 Clay: that would be the easiest way, yes. Setting up the OM env w/ MokoMakefile can be a real PITA, especially for newbies ;) May 06 20:24:49 *w/o May 06 20:24:55 Clay, if you get stuck check http://wiki.openmoko.org/wiki/Talk:MokoMakefile May 06 20:25:01 The silver/black openmoko on openmoko.org is with with that colour scheme going to be available? May 06 20:25:27 alright, thanx, im going to install linux now May 06 20:25:43 Mukunda: you might wanna rephrase that question ;) May 06 20:25:46 Mukunda: note that openmoko is the software May 06 20:25:53 CoreDump|home: sorry. May 06 20:25:54 Yeah. May 06 20:25:58 Just woke up. May 06 20:26:04 The neo1973 I mean. May 06 20:26:05 :-) May 06 20:26:24 The silver/black casing does indeed exist May 06 20:26:38 Most P0 devices were black/silver AFAIK May 06 20:26:41 Ah awesome, tis sexy compared to the photos I've seen. May 06 20:27:12 it looks hot =) May 06 20:27:24 and feels interesting May 06 20:28:13 Hmnse well, hopefully they announce what the discount will be for P1 buyers, soon. May 06 20:28:38 there will be a discount? May 06 20:28:54 From what I heard, the P1 phase is basically right around the corner May 06 20:29:01 oh, a discount for P2 for P1 buyers you mean? May 06 20:29:09 balrog-kun: ah may have been. May 06 20:37:05 Do you mean P1.5? May 06 20:37:14 Not heard about possible P2 discount. May 06 20:37:58 is any kind of discounts planned then? :p May 06 20:38:21 P1+ or whatever the refresh is called. May 06 20:39:31 Hm can't find this on the faq, what kind of cpu is in the neo1973? May 06 20:39:34 balrog-kun: it says so in the wiki with refs to mailing list May 06 20:39:47 Look on the neo1973 page on teh wiki May 06 20:39:59 or on teh hardware information page from the main page May 06 20:40:09 sc3410? May 06 20:40:12 something lie that. May 06 20:40:40 S3C2410A May 06 20:41:00 the hardware refresh will most likely use s3c2440 May 06 20:41:18 Oh hm that seems like a decent enough cpu. May 06 20:41:30 koen: afaik think that's GTA02 May 06 20:41:35 s/think// May 06 20:41:36 balrog-kun meant: koen: afaik that's GTA02 May 06 20:42:11 balrog-kun: could be, dunno about the numbers May 06 20:42:24 * koen is still hoping for omaps in future neos May 06 20:42:27 or blackfins May 06 20:42:31 anything with a DSP :) May 06 20:43:04 'night all May 06 20:43:06 * balrog-kun wonders what's decent enough about s3c2410 :) May 06 20:43:10 omaps would be great May 06 20:46:45 IMO, the s3c2410 has about enough CPU power. May 06 20:46:52 I want less battery use. May 06 20:47:04 Or USB2 May 06 21:01:09 * CoreDump|home wants good video acceleration May 06 21:01:44 * pavelm wants working GPS May 06 21:01:48 * pavelm wants usable display May 06 21:02:12 * pavelm wants at least arrow buttons, and 2 functional keys May 06 21:02:38 pavelm: yeah, a "joypad" would be nice May 06 21:02:53 Decoded GPS packet 0: it contains time in miliseconds. May 06 21:03:13 coredump: I believe joypad is pretty much mandatory... but I guess it would look uncool? May 06 21:04:16 pavelm: well, I trust in FIC's design dept to make it look decent ;) May 06 21:04:30 pavelm: There are drivers for the Wiimote in the kernel ;) May 06 21:04:40 and the Wiimote is bluetooth May 06 21:04:42 pavelm: did you manage to turn on your GPS? May 06 21:04:52 kensan: Ok, I guess I _can_ just ssh from my zaurus :-) May 06 21:05:07 kensan: but then, I could just put gsm card into my zaurus, right? May 06 21:05:21 coredump: I did not try, yet. Laf0rge warned me there are still bugs there. May 06 21:05:22 exactly May 06 21:06:11 w/ a working power-on, our reverse-engineers could have a go at the hammer head May 06 21:09:13 pavelm: do you mean packets with ID 0? May 06 21:09:21 From the GPS? May 06 21:10:24 Hmm. May 06 21:10:30 http://www.globallocate.com/ says hammerhead II will work indoors... May 06 21:10:32 It won't be an absolute time of course. May 06 21:11:04 Speed: Yes, packet 0 _from_ gps looks like milisecond precission timestamp. May 06 21:11:40 I'd expect it to be a time, based on the VCXO clock, showing May 06 21:11:44 will the satellites have to be present in your room for that to work? May 06 21:11:49 skew of each satellite signal. May 06 21:11:58 balrog-kun: No,. May 06 21:12:14 Because of the sensitivity and the fast lock time, it can use signals that others can't. May 06 21:12:28 speed: Yes, I'd expect something like that. May 06 21:12:30 they could add a free bonus satellite to every chip May 06 21:12:33 The hammerhead really is quite a lot more sensitive fundamentally. May 06 21:12:42 speed: (No idea what VCXO is, anyway). May 06 21:12:44 Than your average sirfIII/... May 06 21:12:56 VCXO Voltage controlled crystal oscillator. May 06 21:13:34 This is the precision timebase used for the precise timings of satellite motion. May 06 21:13:35 actually I'd expect it to be time represented in oscilations of incoming signal, or something. May 06 21:13:48 No, as you don't have a reliable incoming signal. May 06 21:13:59 You've got to express it generally in terms of an internal clock. May 06 21:14:35 Which the VCXO is. May 06 21:15:08 Maybe it is not a timestampt, but something that changes with time. Like ammount of bits received from particular sattelite... packets 0 come in groups, and at the beggining they have all the same "timestamp". At the end of the log, they differ. May 06 21:15:30 Maybe. May 06 21:16:17 They state that there is no hard real-time needed by the host. May 06 21:16:30 So the host doesn't need to do precise timing on the serial inputs. May 06 21:16:46 And you couldn't time serial events that finely anyway. May 06 21:16:47 ...but they do not seem to have timestamps on _all_ packets. May 06 21:17:21 Now if packet 0 has contains very precise timestamps for each sattelite, we are done, right? May 06 21:17:28 It can't. May 06 21:17:34 Unfortunately. May 06 21:17:57 The time is not broadcast sub-second. May 06 21:17:58 I mean... not timestamps... the reverse value. May 06 21:18:15 The hammerhead can get a position in under a second. May 06 21:18:43 I mean... not timestamp... how is it called. May 06 21:18:54 Therefore, in at least one mode, it's not using the broadcast time (with 1s granularity reprated at most every 6s) May 06 21:18:58 Number of ticks since start of packet on each carrier. May 06 21:19:48 well, it has to interpolate between broadcast time, obviously. May 06 21:20:00 ...and they may not know precise time within 1 second of cold start, right? May 06 21:20:22 Yes. May 06 21:20:35 gps tutorial: http://www.colorado.edu/geography/gcraft/notes/gps/gps_f.html May 06 21:20:35 I assume that's not actually true cold start. May 06 21:20:59 I assume it's a mode in which you know the actual time to within 10ms. May 06 21:21:12 This way you can ignore the actual contents of the data. May 06 21:21:17 And just sync to the bit edges. May 06 21:21:32 Yep. May 06 21:21:48 You can do this with normal low-quality clocks as long as you've got teh last update within a couple of minutes. May 06 21:22:55 With more smarts, you can with a second or so of data knowing what the satellites should be broadcasting sync up even if you don't hit a packet with the time in immediately. May 06 21:23:20 If you're within say 100ms. May 06 21:23:54 There are - as I see it - two options. May 06 21:24:29 Either the GPS is nice, and you get data + doppler shift + phase delay relative to precise clock. May 06 21:24:40 And this is all wrapped into a nice packet for you. May 06 21:24:49 for each satellite. May 06 21:24:49 Or. May 06 21:25:13 what's the web-browser they've got running on the Neo1973 at the moment? May 06 21:25:22 Is it just lynx/links2? May 06 21:25:23 ...what kind of data? May 06 21:25:24 You have to basically do successive approximation on the raw data, setting up parameters different ways. May 06 21:25:37 I mean data bits the satellite is sending. May 06 21:25:46 I'd expect doppler shift + phase delay in one packet. May 06 21:25:57 ...and the bits in another one... because that's variable length. May 06 21:26:02 That would be nice. May 06 21:26:43 I thought "data" packets (decoded 50bps stuff) should be easy to locate, but it is not :-( May 06 21:26:46 It's possible that the hardware simply extracts the bits, and you have to do successive approximation or something to extract the doppler shift. May 06 21:27:16 Oh, and while I'm away, if anyone has a suggestion for a name of a hacker oriented shop (I and a friend are hoping to open one), then please say my name followed by the suggestion. May 06 21:27:21 I thought extracting the doppler shift is pretty low-level operation. Something you have to do to get the data... May 06 21:27:26 And by extract the bits, I mean 'reports which bins of 1024 in the specified time have over x energy'. May 06 21:27:38 Unfortunately not. May 06 21:27:57 You can actually decode the data generally without knowing doppler. May 06 21:28:05 ...and something hw needs to do _unless_ you want to communicate with sw heavilly. May 06 21:28:36 The max delta-v is something like 3Km/s. May 06 21:28:39 speed: Is not basic "get the lock" loop (per channel) something like "for all dopplers and all phases, try decoding"? May 06 21:28:50 No. May 06 21:29:01 speed: which has to be redone each milisecond or so? May 06 21:29:11 The maximum difference in speed between you and the satellite is something around +-2Km/s. May 06 21:29:34 This is about a part in 75000, compared with light speed. May 06 21:30:04 The hammerhead is a bit special. May 06 21:30:17 It does not 'lock' as such. May 06 21:30:49 It has 1024 correlators that try to evaluate the energy present in each possible delay 'bin'. May 06 21:31:19 so, you instantly start to recieve a signal if it's there. May 06 21:31:25 speed: And hh differs from "normal" gps, because normal gps has only 2 correlators per channel? May 06 21:31:34 Pretty much. May 06 21:31:55 Which have to do the 'slide along the waveform looking for lock'. thing. As illustrated in the colorado website. May 06 21:32:00 Hmm. Ok. Thats pretty much 'more powerful' May 06 21:32:22 But I'd expect 1024-byte long packets somewhere :-) May 06 21:32:32 Or 1024 nibble/... May 06 21:32:42 that's still too long. May 06 21:32:54 Or it's reasonable that there is a 'report bins in which energy >100' May 06 21:33:03 biggest we have here is something like 64. May 06 21:33:03 Which would be much shorter. May 06 21:33:20 ...maybe it can be 2 bits per bin or something? May 06 21:33:27 Possible. May 06 21:33:35 ..no, that does not fit, still. May 06 21:33:53 What would help somewhat is more data. May 06 21:33:56 ...and this info would have to be broadcast before each navigation message, no? May 06 21:34:07 We'll get more data soon enough. May 06 21:34:31 I'm thinking of something like 'GPS wrapped in foil' would be handy for nailing down satellite signals. May 06 21:34:40 I'd also like to do "active attack" -- insert my soft between hammerhead and their sw, modify bitstream, and look what happens. May 06 21:34:53 That'd be fun. May 06 21:34:58 * balrog-kun hopes to get his hands on a TomTom GO 910 one of these days May 06 21:35:00 ...you may need lots of foil for hammerhead -- if they can do indoors. May 06 21:35:13 Did you see the filing cabinet one? May 06 21:35:58 speed: I heard it works inside metal box, but I did not see photos. May 06 21:36:10 Is tomtom 910 really that much better than other GPSes? May 06 21:36:27 http://www.globallocate.com/SEMICONDUCTORS/SEMI_LIBRARY_Frameset.htm May 06 21:36:29 (script) wiki RecentChanges: [[Talk:Hammerhead_protocol]] [[Hammerhead_protocol]] May 06 21:36:31 is interesting. May 06 21:36:37 pavelm: i don't know, but it uses a Hammerhead afaik May 06 21:37:00 They have some very impressive graphs in the pdfs on those. May 06 21:37:22 And also other PDFs on teh technology are interesting reading. May 06 21:37:41 Giving hints. May 06 21:37:45 Alas no datasheets :/ May 06 21:38:20 speed: Well, it is _their marketing_. Not sure if it can be trusted. May 06 21:38:24 There are also links to potentially interesting patents. Which I haven't read. May 06 21:38:28 True. May 06 21:41:20 Try this for some interesting reading: http://www.gmat.unsw.edu.au/snap/publications/opensourcegps.pdf . Code is GPLed. May 06 21:44:41 Evening guys. May 06 21:44:50 Evening. May 06 21:47:47 That opensourcegps.pdf seems quite interesting. May 06 21:48:11 Though major parts of it are irrelevant - as they are done completely differently, yes, indeed. May 06 21:52:14 speed: Yes, hopefully some parts of opensourcegps are already in hw. May 06 21:54:36 Hi pavelm! May 06 21:55:53 Does the opensource gps conform to the itar stuff? ;) May 06 21:56:06 It doesn't need to. May 06 21:56:22 It's only if you bundle it with hardware and import/export it that it needs to. May 06 21:56:49 Hi, Elrond! May 06 22:01:32 UI screenshots from today for anyone interested http://hentges.net/tmp/screenshots/Neo1973/20070506/ May 06 22:03:11 CoreDump|home - Is this from a fully standard build? May 06 22:03:18 CoreDump|home: interesting. May 06 22:03:30 yes, from a OM build from today May 06 22:03:53 nice shots... but does the dialer actually work? ;-) May 06 22:04:28 pavelm: AFAIK yes May 06 22:04:41 time to try it out heh May 06 22:04:44 Okay, the texts on the Desktop are too small. May 06 22:05:59 core: The "Y ?" icon makes me wonder. (I was testing may-01 build, and both sound and gsm were broken). May 06 22:06:48 Elrond: It is indeed small but readable May 06 22:07:00 pavelm: I've inserted a sim and am booting now May 06 22:07:13 CoreDump|home - I have bad eyes. I'd need a microscope. May 06 22:07:35 the theme is far from finished =) May 06 22:07:50 Silly emulation aid. May 06 22:07:52 so there is hope they'll 'up the font May 06 22:08:03 Obtain a rectangle of cardboard 2.8" diagonal. May 06 22:08:23 Hold at normal distance for phone, and back away from screen till same size. May 06 22:08:49 SpeedEvil - ;-) May 06 22:08:58 SpeedEvil - Or print stuff out at 300dpi? ;o) May 06 22:09:10 That takes me back. May 06 22:09:18 * SpeedEvil has played MUDs on a printer. May 06 22:09:25 (when my monitor broke) May 06 22:10:02 Huh?! May 06 22:10:04 speed: Hehe, we did the same with unix shell. But that printer was definitelly not 300dpi :-) May 06 22:10:09 on the dialer, why the small letters on the '0' and '#' buttons? May 06 22:11:14 elrond: Huh what? May 06 22:11:27 pavelm: the dialer segfaults :\ May 06 22:12:06 core: do you at least get signal icon? May 06 22:13:29 pavelm - I'm surprised over the mud-on-printer thing. Even logging into my linux box would have been a challenge ;) Not to mention, I have no printer at home. ;) May 06 22:14:29 Printer certainly helps ;-) May 06 22:14:40 It was a dot matrix one. May 06 22:15:03 And it had a special mode - if it had no data for 1s, it fed the paper up so you could see the last line printed. May 06 22:15:13 Add fanfeed paper, and ... May 06 22:15:24 hehe. May 06 22:15:42 Did it scroll the paper again down, when the data continued for that line? May 06 22:16:23 Yes. May 06 22:16:43 Fun. :) May 06 22:16:54 pavelm: no signal:( May 06 22:17:30 core: Too bad, so gsm part is still not fixed. May 06 22:17:35 CoreDump|home - I heard, that in the Settings Menu, there would be an option to increase the font size... is there one? May 06 22:18:13 speed: Hammerhead seems nice on paper. I'm just afraid that someone will do "ok, so this is super sensitive.... lets use smaller antena, then". May 06 22:18:21 :/ May 06 22:18:40 The antenna size does not seem unreasonable. May 06 22:19:38 "ok, so lets use crappy amplifier" :-) May 06 22:20:11 I think, the hammerhead has the lna integrated. May 06 22:20:43 Elrond: http://hentges.net/tmp/screenshots/Neo1973/20070506/fontsize-00.jpg May 06 22:21:33 Ok... now scale the font to something reasonable.... Like 8 lines per screen. May 06 22:21:37 ...does the UI still work? May 06 22:22:31 this font-size only affects the desktop font IIRC May 06 22:23:16 Elrond: yes, it's integrated. May 06 22:23:37 That's unusable, then. I need to read my sms-es, too, and would like to read the menus... May 06 22:24:07 what was the device with the hammerhead? tomtom 901? May 06 22:24:43 tomtom go 910. Will look at its reviews. May 06 22:28:13 The "pick a font" title is also straight unreadable. May 06 22:28:50 (I can read it slowly with my standard glasses on a 100dpi screen. May 06 22:28:52 ) May 06 22:31:29 Hmm...review claims tomtom go 910 is sirfIII :-(. May 06 22:31:36 Ok, I guess I should get some sleep. May 06 22:32:33 Night. May 06 22:47:41 Does anyone here have the "Nokia Content Publishing Toolkit" thingy for linux? The Nokia idiots removed the linux version May 06 23:05:33 re May 06 23:12:30 anyone here know how to get a ubuntu compiler working? May 06 23:16:50 im really new to programming and i jst got ubuntu May 06 23:17:59 Clay - What compiler do you need? May 06 23:18:44 i dont know, i want to start programming for the OpenMoko project May 06 23:32:02 can anyone help? May 06 23:33:43 Hm... I think you could _really_ use a beginner's guide :) May 06 23:33:51 Any preference for a particular language? May 06 23:34:11 The core apps are all C May 06 23:34:27 C with GTK, which gives it an entirely different dimension :) May 06 23:34:45 i dont really have a preference for programming language May 06 23:35:26 Clay: there's an online book "Dive Into Python" that you can access May 06 23:35:53 type 'python' on your linux box to check if it's installed May 06 23:36:06 Is Python the right choice though? May 06 23:36:23 what are you suggesting? VB? May 06 23:37:01 Clay: this will be an extrodinary learning experience for you May 06 23:37:17 C :D May 06 23:37:19 no matter what language you begin with, it will be time well spent May 06 23:37:26 I guess so. May 06 23:37:31 AS LONG... as it's not VB. May 06 23:37:37 yuk May 06 23:37:42 yeah it probably will May 06 23:38:23 you'll need to become comfortable with programming on your linux box before you write phone apps May 06 23:39:22 Python is a great language. I love it personally, but if you are wanting to shift to C later it is a bit different of a syntax. May 06 23:39:50 alright May 06 23:40:24 Clay: another alternative would be Java--there is a ton of tutorials at http://java.sun.com May 06 23:40:54 it is my personal opinion that C has too many guns and knives for a beginning programmer May 06 23:41:35 you have to know gdb and make, very nontrivial apps May 06 23:41:50 Java is a beautiful language. May 06 23:41:52 and you have to know how to invoke gcc with the proper options May 06 23:42:09 I'd have to agree (coming from someone who can follow simple C but not much else). May 06 23:42:11 Java has syntax much like the C syntax May 06 23:42:56 More or less. May 06 23:43:16 If you... ignore the part where it's Object-Oriented :) May 06 23:43:22 Clay: if your Linux box is up and running then you're ready to learn. May 06 23:43:31 C++ ? May 06 23:44:55 Not quite, but I guess there are similarities. May 06 23:45:26 so what do u guys think i should go with? May 06 23:45:36 Clay: you'll also want to go to B&N and purchase "Vi IMproved--VIM" by Steve Oualline. You'll need to know how to use an editor. May 06 23:46:19 heh :) May 06 23:46:23 lol May 06 23:47:36 morning May 06 23:48:17 morning May 06 23:48:38 Anyone here had experience with the Nokia N800? May 06 23:48:38 rwhitby-airport: 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 06 23:49:07 heh, ~botsnack May 06 23:50:02 lol May 06 23:50:21 rwhitby-airport, my friend has, I've only tapped it a few times :) **** ENDING LOGGING AT Mon May 07 02:59:56 2007