**** BEGIN LOGGING AT Fri Feb 01 02:59:57 2008 Feb 01 05:18:00 * * OM Bug 1211 has been created by erin_yueh(AT)openmoko.com Feb 01 05:18:01 * * phone screen freeze in some UI applications Feb 01 05:18:02 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1211 Feb 01 05:35:44 first command I tried inside my testing framework, AT+COPS=? actually has a buggy parsing method in gsmd. Feb 01 05:36:01 this is going to be so useful. Feb 01 08:03:55 Cool Dating Club site ===> new-datingclub eu Feb 01 08:06:28 good morning Feb 01 08:17:54 Hopscotch, good morning Feb 01 08:48:39 cesarb: if you need somewhere to put your image, I'll host it for you Feb 01 08:50:37 with debug board one can step-through u-boot as well as Linux code through JTAG? Feb 01 08:50:54 jtag lets you step through anything Feb 01 08:51:02 (though not into mordor) Feb 01 08:52:23 SpeedEvil, just wondering what would be the price of freerunner, advanced version Feb 01 08:52:38 $6000 is reported Feb 01 08:52:40 err Feb 01 08:52:43 $600 Feb 01 08:52:50 SpeedEvil, fair enough Feb 01 08:52:57 I wouldn't be astonished if that price changes Feb 01 08:56:07 SpeedEvil, but, wiki has a list of people interested in buying gta02? similar to the first list Feb 01 08:57:04 So? Feb 01 08:57:32 The wiki has a list of lots of stuff. Feb 01 08:57:36 SpeedEvil, found it, http://wiki.openmoko.org/wiki/Buying_Interest_List , Feb 01 08:57:39 very, very little of it is official Feb 01 08:57:48 SpeedEvil, just to see how many people are interested in gta02 Feb 01 09:02:21 mbuf: JTAG lets you step through anything, but you don't want to use it to debug application code Feb 01 09:02:34 mbuf: it's fine for debugging kernel code though Feb 01 09:02:34 drath, ok Feb 01 09:02:57 mbuf: for application, gdbserver is a lot easier to use, and written especially for that purpose Feb 01 09:02:57 drath, how about u-boot with debug board? Feb 01 09:03:10 mbuf: yeah, you can step through u-boot, too Feb 01 09:03:16 drath, ok Feb 01 09:03:24 mbuf: you should be able to step through it from the very first instruction (0x0) Feb 01 10:10:13 openmoko: 03chris * r4008 10/trunk/src/target/OM-2007.2/applications/openmoko-dialer2/ (ChangeLog src/phone-kit/moko-network.c): Feb 01 10:10:13 openmoko: * src/phone-kit/moko-network.c: (io_func): Feb 01 10:10:14 openmoko: Handle all IO conditions, bail on unknown conditions (noticed problems Feb 01 10:10:14 openmoko: when doing otherwise) Feb 01 10:12:16 morning Feb 01 10:14:45 Mornin Feb 01 10:27:58 ScaredyCat: thanks, but it's too alpha-version so far for it to be available to the general public Feb 01 10:28:27 well, i wan't suggesting everyone would be able to dl it :) Feb 01 10:28:43 the offer is there if you need it Feb 01 10:31:03 cesarb: what cool shit are you cooking? Feb 01 10:31:50 ScaredyCat: I'd probably want to make it available to the public when it's more near ready (most important frequency-depentend drivers working) Feb 01 10:31:57 XorA: cpufreq Feb 01 10:32:07 cesarb: sweet Feb 01 10:32:12 http://www.linuxdevices.com/news/NS9710143183.html interesting =] Feb 01 10:32:21 k, np - just shout if you want it Feb 01 10:33:02 ScaredyCat: thx Feb 01 10:33:17 XorA: http://wiki.openmoko.org/wiki/User:CesarB/cpufreq Feb 01 10:35:13 XorA: (that page needs a bit of an update, btw; I've changed several things in the code already) Feb 01 10:35:21 cesarb: you audio problems with cpufreq dont apply to gta02 Feb 01 10:35:57 XorA: well, the code so far doesn't even run on gta02 ;-) Feb 01 10:36:07 cesarb: hopefully soon :-) Feb 01 10:36:15 XorA: and gta01 is what I have and I'm focusing on... Feb 01 10:36:23 cesarb: but 2442 cpu can send the crystal straight to codec without scaling Feb 01 10:36:29 XorA: cool Feb 01 10:36:34 cesarb: and audio codec is always clock master for audio Feb 01 10:37:25 XorA: in fact, I didn't look at the IIS block so closely yet (I've only done timer4, nand, serial, and framebuffer so far, and I'm looking at usb device) Feb 01 10:37:54 XorA: there's one problem in porting my code to the gta02: the frequency table I found so far for it is very limited Feb 01 10:38:03 cesarb: well gimme a shout if you need help with audio part Feb 01 10:39:08 XorA: thanks Feb 01 10:39:31 its fairly self explanatory though, the codec dividers have helper functions Feb 01 10:39:34 * cesarb suddenly notices the cpu listed on the [[Neo1973 GTA02 Hardware]] page changed yesterday Feb 01 10:40:34 cesarb: from/to? Feb 01 10:42:53 Core 2 duo Feb 01 10:44:17 Stephmw: http://wiki.openmoko.org/index.php?title=Neo1973_GTA02_Hardware&diff=23400&oldid=23183 Feb 01 10:54:18 cesarb: that's not a change in the CPU? Feb 01 10:54:34 oh - right Feb 01 10:54:39 * SpeedEvil scrolls to the right Feb 01 10:55:13 SpeedEvil: it just adds a small extra designator, which might not mean much, or might mean a completely different architecture. With hardware, there's no way to know. Feb 01 10:55:14 IIRC that simply specifies a package Feb 01 10:56:54 SpeedEvil: it might also specify a speed grade Feb 01 10:57:14 XorA: did you fux with kaiser yet? Feb 01 10:58:22 dcordes: no, Im not brave Feb 01 11:00:10 cesarb: odd - the datasheet I have has no mention of part code tables Feb 01 11:01:19 SpeedEvil: it's probably in a different pdf (one about ordering options) Feb 01 11:01:24 * cesarb has to go now Feb 01 11:01:42 wave. Feb 01 11:05:08 XorA: you don't want to break it? Feb 01 11:13:02 Core 2 Duo ? Feb 01 11:13:03 LOL Feb 01 11:17:22 where oh where is my Dark Heresy book Feb 01 12:43:44 are there any qemu-neo1973 users on PPC (or anything non-x86)? Feb 01 12:49:31 I vaguely recall others asking the same question and getting a no Feb 01 12:50:23 (I tried it on amd64!) Feb 01 12:50:26 i would doubt it Feb 01 12:50:35 there arent many desktops running on x86 anymore Feb 01 12:50:40 on non-x86 Feb 01 12:52:25 good.. a new code generator was just merged upstream that removes dependence on gcc (ie. gcc 3) but it only works for x86 and amd64 hosts for the moment Feb 01 12:55:03 * balrog-kun will proceed to update qemu-neo1973 some time too... Feb 01 13:05:13 * SpeedEvil ponders ms+yahoo Feb 01 13:05:26 and boggles at $44bn. Feb 01 13:09:02 * buz thought the same Feb 01 13:09:12 who still uses yahoo anyway Feb 01 13:09:43 buz: flickr might be part of the bargain Feb 01 13:09:56 buz: and if that happens, my photos and I are moving elsewhere Feb 01 13:10:38 first the ftc needs to agree Feb 01 13:10:59 and the eu may yet rain on that parade Feb 01 13:13:12 why does my serial-terminal crap out a shitload of ^ on me? Feb 01 13:58:55 someone knows pyroute? Feb 01 13:58:56 [eichi@eichi-laptop pyroute]$ python gui.py Feb 01 13:58:56 Downloading POIs from OSM Feb 01 13:58:56 and then, noting happens Feb 01 13:59:23 no error, no exit Feb 01 13:59:32 eichi: same here ;) Feb 01 13:59:47 hdr, wellcome ;) Feb 01 14:06:16 hdr, i got it Feb 01 14:06:42 try downloading to pyroute/data/poi_all.osm from http://www.informationfreeway.org/api/0.5/node Feb 01 14:06:46 seems to be a download bug Feb 01 14:07:00 eichi: thx ;) Feb 01 14:09:04 hdr, but it need more files Feb 01 14:09:08 doenst route at all Feb 01 14:10:08 Loading OSM data from data/routing.osm Feb 01 14:10:08 No such data file data/routing.osm Feb 01 14:10:36 eichi: yeah i see only a straigt line between start end destination Feb 01 14:11:04 same here Feb 01 14:11:09 we need a routing.osm Feb 01 14:12:19 url = "http://informationfreeway.org/api/0.5/way[bbox=%1.4f,%1.4f,%1.4f,%1.4f][highway|railway|waterway=*]" % (lon-size,lat-size,lon+size,lat+size) Feb 01 14:12:30 seems to big for downloading manually Feb 01 14:13:30 openmoko: 03werner * r4009 10/branches/src/target/kernel/2.6.24.x/patches/ (4 files): (log message trimmed) Feb 01 14:13:30 openmoko: clean-fiq.patch Feb 01 14:13:30 openmoko: Change ipc struct type from _fiq_ipc to fiq_ipc Feb 01 14:13:30 openmoko: From: Andy Green Feb 01 14:13:30 openmoko: Signed-off-by: Andy Green Feb 01 14:13:31 openmoko: smedia-glamo.patch: Feb 01 14:13:33 openmoko: - moved complaint about include/asm-arm/arch-s3c2410/irqs.h here Feb 01 14:26:17 eichi and hdr, Iøve had the same "problem" before also. You need to set your location on the map, and then enter the menu and somehow tell the program to download the routes Feb 01 14:26:27 *I've Feb 01 14:29:51 Ebbe, ah okay Feb 01 14:32:26 Hellow fellow ignorant Pedophiles, have any child porn to share today? Feb 01 14:33:28 Hellow fellow ignorant Pedophiles, have any child porn to share today? Feb 01 14:33:30 I think you're on the wrong channel Feb 01 14:33:57 #fbi is thataways Feb 01 14:35:50 * buz shakes head in disbelief Feb 01 14:36:24 * Ebbe just found it funny ;) Feb 01 14:42:56 i think that even asking for child porno should be illeagle Feb 01 14:43:38 including the fbi Feb 01 15:39:23 eichi: do you know where the data.osm should go? Feb 01 16:04:37 zilux: Hey, thanks so much for the navit stuff yesterday! Feb 01 16:04:51 zilux: I somehow had the 0.0.1 ipk that I was installing :-/ Feb 01 16:05:56 DYNAMIC ENTRY Feb 01 16:11:58 wibbit, data/poi.osm Feb 01 16:16:32 eichi: How big is the poi_all.osm? Feb 01 16:17:13 [eichi@eichi-laptop ~]$ du -h test/pyroute/data/poi_all.osm Feb 01 16:17:13 4.0K test/pyroute/data/poi_all.osm Feb 01 16:17:47 zilux: Anyway, it all works now Feb 01 16:17:59 du -sh poi_all.osm Feb 01 16:17:59 5.2M poi_all.osm Feb 01 16:23:05 eichi: All working now :) Feb 01 16:23:47 wibbit, nice Feb 01 16:24:16 eichi: Do you know if any one is working on a client to do on the fly editing of the map? Feb 01 16:27:12 wibbit, I dont know. I would like this to. I use a N800 for tracking für OSM and Josm on laptop for mapping. In the fly would be perfect, it will be the futur, but the time has not come ;D Feb 01 16:27:56 eichi: It would certainly be handy to be able to spend a little bit of time, and uploading the GPS info and initial mapping data, Feb 01 16:28:43 eichi: Though saying that, I've just noticed that there is a "sketch" mode, which allows you to draw on the map, which could be handy Feb 01 16:30:47 yeah, we thought about the idea, that you can mark things on map, which are maybe not right and upload this marker, so someone can see all "problems" in this homelands and can correct them Feb 01 16:31:12 is this, what you mean? Feb 01 16:42:38 wibbit: Sorry to jump in late here, but may I ask what mapping app you're talking about? Feb 01 16:42:51 Is there still uncertainty about whether the GTA02 will support 850MHz? Feb 01 16:43:25 wurp2: No problem, never too late :), pyroute Feb 01 16:43:35 malsyned: I don't know, but I know when I ordered my 01 and asked about 850 on it someone internal to OpenMoko told me the 02 will have 850 Feb 01 16:43:44 Ah, I would love to get that working on the neo Feb 01 16:43:47 it is just too slow Feb 01 16:43:59 I need to profile it and see what can be done about it Feb 01 16:44:02 wurp2: awesome. thanks. Feb 01 16:44:24 eichi: Well I certainly like the idea of being able to upload a "temp image" or draft, which could then be edited with JOSM at home in detail, but being able to enter road names, or post boxes, and the ilk. Feb 01 16:44:25 I'd been thinking I would port pypy or psyco, but on further thought I should really figure out what the problem is before I try to solve it :-) Feb 01 16:44:34 malsyned: You're welcome :-) Feb 01 16:45:48 It would be cool to be able to sketch a road name on the map, with an arrow to the road (said road perhaps being created on the fly from gps data) and then hide the sketch Feb 01 16:46:13 Later on a machine with a keyboard you could convert the sketch into an OSM compatible road name Feb 01 16:46:33 malsyned: 850 has been demonstrated on modified 01s Feb 01 16:46:52 malsyned: I dunno if there are reliability problems, etc. It's still very early in the 850 testing Feb 01 16:48:05 wurp2: sure. If it doesn't work at 850, it's not very useful to an American like myself, is it? Feb 01 16:48:25 (I'm not really up on this whole frequency issue) Feb 01 16:50:42 malsyned: I'm in the US (Dallas/Fort Worth area) and it works fine for me Feb 01 16:50:53 There are big swaths of the US that only have 850 coverage Feb 01 16:51:03 but almost all metroplexes have 1900 too Feb 01 16:51:25 If you scout around on wiki.openmoko.org there's a link to a 1900 coverage map somewhere, I think Feb 01 16:51:52 I should say, instead of big swaths, almost all of the US is 850 coverage, by land area Feb 01 16:52:03 but I think by population most people are in a 1900 area most of the time Feb 01 16:52:20 we also need a highhiking symbol for OSM Feb 01 16:54:50 * Kero has no problem with 1900-only where he is (litle town along the Hudson river). But then, for other reasons, had to buy a $40 dual 850/1900 phone anyway... Feb 01 16:55:10 also note, I'm only in the USA for a few months Feb 01 16:55:36 kdean06: you were interested in my (lib)gsmd testing framework, no? Feb 01 16:55:49 Yeah. Feb 01 16:56:03 I have the roundtrip in the code, reasonably nice for testing itself. Feb 01 16:56:20 expect a release sometime this weekend Feb 01 16:56:26 Awesome. :) Feb 01 16:56:54 testing can be done on a PC w/ linux; hope you like Ruby, too :) Feb 01 16:57:44 Kero: Is there Ruby for the neo? Feb 01 16:57:45 www.mirclan.net www.mircmerkezi.net www.sohbet-turk.org http://mirc.sohbet-turk.org http://mirc.mircmerkezi.net http://haber.mirclan.net Feb 01 16:57:48 www.mirclan.net www.mircmerkezi.net www.sohbet-turk.org http://mirc.sohbet-turk.org http://mirc.mircmerkezi.net http://haber.mirclan.net Feb 01 16:57:50 www.mirclan.net www.mircmerkezi.net www.sohbet-turk.org http://mirc.sohbet-turk.org http://mirc.mircmerkezi.net http://haber.mirclan.net Feb 01 16:57:56 I sensed that... Feb 01 16:58:01 me too Feb 01 16:58:08 wurp2: Or as opposed to sketching, I suppose things like temp POI's could be used for that. associate a POI with the road, and put in the name, possibly. Feb 01 16:58:47 wibbit: True, but e.g. if I'm driving down the road, even if I pull over to do it, sketching is much faster than typing on a neo-type device Feb 01 16:59:18 wurp2: May be for you, I think I'd find tapping out the road name, a lot quicker then trying to sketch it Feb 01 16:59:20 wurp2: certainly, compiled it myself :) Feb 01 16:59:23 wurp2: my hand writing sucks Feb 01 16:59:47 http://chmeee.dyndns.org/om/ruby/feed/ Feb 01 17:00:37 including gtk2 bindings Feb 01 17:01:19 wibbit: Mine too, but it's still much quicker for me than typing with a stylus Feb 01 17:01:24 Kero: Cool! Feb 01 17:01:29 * Kero needs a TODO list for all the pending stuff he has... now that he has time to work on openmoko Feb 01 17:01:37 wurp2: okie Feb 01 17:03:44 wibbit: Have you guys tried pyroute on the neo? Feb 01 17:04:17 It takes a LOOOONG time to start up, and then when you navigate to any of the menus, as far as I can tell it never processes any further clicks Feb 01 17:04:25 I don't have a neo, was just trying it on my desktop Feb 01 17:04:34 ah Feb 01 17:04:51 I assume it is fixable. I've just had higher priority things to work on Feb 01 17:05:25 And navit works fine, although I don't like the UI as much, I'm not as prone to add to a C app as python, and there are a few features in pyroute not in navit Feb 01 17:05:29 e.g. sketching Feb 01 17:30:06 I know OM mentored some people last year for gsoc, will OM do that again this year? Or will it probably be left to OE? Feb 01 18:28:48 hi Feb 01 18:31:33 Anyone knows how to modify the kernel configuration before to build openmoko? Feb 01 18:32:33 There is the file "moko/build/tmp/work/neo1973-angstrom-linux-gnueabi/linux-openmoko-2.6.22.5-moko11+svnr3801-r13/linux-2.6.22/.config" that has the configuration but it is overwritten when I rebuild the kernel Feb 01 18:36:12 jmichel: how do you build the kernel? Feb 01 18:42:27 mwester: with "make rebuild-package-linux-openmoko" Feb 01 18:45:21 mwester: and I removed the tag that deletes the source after the build so I could make some changes... but the "make rebuild-package-linux-openmoko" overwrites my changed woth the original version Feb 01 18:46:07 Yep. That's the way it's supposed to work -- at that high a level, the concept of build targets and dependencies is at the level of the entire kernel, or an entire package. Feb 01 18:46:37 You need to work a bit lower in the build structure - bitbake. Feb 01 18:48:05 Basically what I do is to first build the kernel, then edit the sources or update the defconfig - whatever is necessary - just as you've been doing. Feb 01 18:49:39 Is this folder the path of the kernel: "openembedded/packages/linux/linux-openmoko/" ? Feb 01 18:49:52 Then delete the "do_compile" stamp for the kernel -- you'll find that in the buld/tmp/stamps/neo1973-angstrom-linux-gnueabi directory, named linux-openmoko-2.6.22.5-do_compile Feb 01 18:50:39 No, the openembedded/* tree is the origin of the build recipes -- that's where patches and config info would go when a modification is complete and is being submitted. Feb 01 18:51:01 ok Feb 01 18:51:53 Once you've deleted the appropriate do_compile stamp, you can then build with the generic target, and the bitbake system will see that only the compile step is missing for the kernel, so it won't unpack a new kernel and re-apply the patches, etc. Feb 01 18:53:32 ok I'll try this, thanks for the hint Feb 01 18:54:12 Another way - do "bitbake -c clean linux-openmoko" then "bitbake -c patch linux-openmoko". Edit the files, then "bitbake linux-openmoko" to run the rest of the build steps. Feb 01 18:55:14 ( source the setup-env file before you run the bitbake commands) Feb 01 18:55:40 Yes, and run the bitbake command while in the "build" directory. Feb 01 19:26:27 does somebody have stats how the voltage drop of the battery stands in relation to the charge-status? Feb 01 19:31:42 thomasg: You mean how much battery time is left? Feb 01 19:31:51 I've seen some graphs, but not lately Feb 01 19:31:57 It's not linear for sure Feb 01 19:32:24 CM, yeah, I want to know when it's over :) Feb 01 19:32:40 neo is running for about 13h now (mostly in suspend) and at 3.785V Feb 01 19:33:08 josch made some graphs, but I can't remember the url Feb 01 19:33:31 thomasg: it's standard lithium-ion Feb 01 19:33:32 same for me :) Feb 01 19:33:38 SpeedEvil, yes Feb 01 19:33:43 but what's standard? :) Feb 01 19:34:00 thomasg: ballpark it's 'linear' between 4.1V and 3.3V or so. Feb 01 19:34:15 with that being about 95% to 5% Feb 01 19:34:31 I need to dig out and fix my patch for that. Feb 01 19:34:36 ah, that's what I wanted to know. thanks Feb 01 19:35:46 however - looks the standbytime is getting at least a bit better Feb 01 19:36:20 using th clock ? Feb 01 19:37:46 13h is still way too poor for a phone, imho Feb 01 19:37:49 How are you suspending? Feb 01 19:38:03 it'll get >>13h Feb 01 19:38:07 baze: True, but this is just an early stage of a work in progress Feb 01 19:38:17 to be honest, i'd rather settle for 10x of that to be satisfied... Feb 01 19:38:52 my old samsung had to be recharged like every 3 days which really sucked Feb 01 19:38:54 Even my crappy little phones w/ nothing on them barely last 3 days... Feb 01 19:39:01 in my opinion Feb 01 19:39:03 http://www.reghardware.co.uk/2008/02/01/sandisk_prices_32gb_sdhc/ Feb 01 19:39:09 Keep an extra battery :-) Feb 01 19:39:20 For 20-30 hours, low-power modes are plausible Feb 01 19:39:29 Just wish it was a micro.. Feb 01 19:39:33 for 100-200 hours, we need GSM fixes Feb 01 19:39:36 * XorA|gone has had the neo working for 24 hours Feb 01 19:39:54 and suspend-RAM working well Feb 01 19:39:57 Can anyone tell me how they're doing suspend/other low power modes? Feb 01 19:40:02 I've never owned a phone that lasted longer than 3 days on a single full charge so I'd be happy if it met that. Feb 01 19:40:10 wurp2: hacked kernels and stuff Feb 01 19:40:17 SpeedEvil: do you think this will be possible (not by tomorrow or the next month, but maybe due summer or so) Feb 01 19:40:17 wurp2: it's not user-ready Feb 01 19:40:23 SpeedEvil: ah, thanks Feb 01 19:40:28 bart___: next month maybe Feb 01 19:40:46 bart___: o a limited extent anyway. Feb 01 19:40:52 SpeedEvil: You think this stuff might roll into the OM mainline kernel by next month? Feb 01 19:40:56 No Feb 01 19:40:57 ok, so there still _is_ potential Feb 01 19:41:10 there may be custom patches and kernels available in a monthish Feb 01 19:41:12 that's at least a good thing to hear Feb 01 19:41:23 What is the gsm standby power consumption? Feb 01 19:41:25 I'd not expect it to be quite stable perhaps for a little bit more Feb 01 19:41:39 wurp2: it depends. At the moment it's 1/40th of battery/hour Feb 01 19:41:39 SpeedEvil: Cool, thanks Feb 01 19:42:02 So like 30 mA? Feb 01 19:42:03 wurp2: It can reportedly be gotten - with firmware - down to 1/100 -1/200th Feb 01 19:42:10 wurp2: yes, about that Feb 01 19:42:27 That firmware figure would be awesome Feb 01 19:42:42 As kdean06 says, I don't think I've ever had a phone that would make it more than 4 days, tops. Feb 01 19:42:47 CM, yeah I did some graphs - did you already find the url? Feb 01 19:42:47 nice. where do I get new firmware? :) (sending unit elsewhere, iirc) Feb 01 19:42:48 I have had many Feb 01 19:42:53 basic nokias Feb 01 19:43:04 My current one gets IIRC 12 days Feb 01 19:43:05 Kero: I think there is a plan to deliver something to update gsm firmware Feb 01 19:43:09 k800i lasts about 5 days here Feb 01 19:43:13 CM, http://rabenfrost.net/openmoko/battvolt.png http://rabenfrost.net/openmoko/battvolt2.png Feb 01 19:43:14 Kero: License agreements are still outstanding Feb 01 19:43:23 b+W LCDs FTW Feb 01 19:43:26 wurp2: there is, but I haven't heard anything definite on it Feb 01 19:43:37 depends of course on how much you actually make calls Feb 01 19:43:48 But for now, you have to send it to Michael Shiloh (for the SIM update; presume it would be the same for the power one) Feb 01 19:43:49 josch: Ah, thanks Feb 01 19:43:59 CM, you're welcome ^^ Feb 01 19:44:10 Kero: Me neither, I was just assuming maybe you hadn't heard about it Feb 01 19:44:15 but for a bit browsing, rss, email and some calls and sms it's about 5 days in the last weeks for me Feb 01 19:44:22 wurp2: no - it's been reported that it should be possible to flash in the fiels Feb 01 19:44:27 field Feb 01 19:44:37 with or without jtag? Feb 01 19:44:39 SpeedEvil: That's what I was talking about re: the license agreements Feb 01 19:44:56 qh Feb 01 19:45:03 SpeedEvil: Afaik you certainly can't flash in the field now Feb 01 19:45:06 * SpeedEvil is also watching TV Feb 01 19:45:13 np Feb 01 19:45:20 TV will rot your brain Feb 01 19:45:24 and we need it for OM kernel updates Feb 01 19:46:14 cesarb is working on much of the cpufreq stuff at the moment which will be a part of the solution. Feb 01 19:47:06 the other part is rapid suspend and resume from RAM Feb 01 19:47:18 question: dialer can use dbus to make a call, no? which daemon/lib/whatever receives the dbus message to pass it to gsmd/firmware ? Feb 01 19:47:33 Ideally so that it can even suspend within a call if you're not using the computer Feb 01 19:50:13 SpeedEvil: that would be very nice; I'd guess about doubling battery lifetime when making calls? Feb 01 19:50:29 Kero: maybe not quite that Feb 01 19:50:41 Kero: but a substantial slice - though less with cpufreq Feb 01 19:56:07 The reports I've heard indicate that the 1700 mAh battery should work in the 01, too Feb 01 19:56:19 Which would add almost 50% to lifetime Feb 01 19:56:31 SpeedEvil: antenna takes so much juice when transmitting? or is power mgmt for the cpu+screen already better than I expect ;) Feb 01 19:56:50 doesn't really matter; would still be very nice! Feb 01 19:56:54 transmitter uses lots of power Feb 01 19:57:45 are batteries already for sale, separately? I'd order one while I'm still in the states. Feb 01 19:58:26 I'm not sure the batteries are available, period Feb 01 19:58:32 (the 1700 mA ones) Feb 01 19:58:54 The reports I'm referring to are the internal OM devs using the 1200 mAh batteries in their FreeRunners Feb 01 19:58:56 hehe; even a spare v01 one (arne't they s/thing like 1500, 1600 mAh anyway?) Feb 01 19:59:18 or am I confusing my old iPAQ battery... Feb 01 19:59:23 as in, I haven't heard any indication that any of the internal devs have the 1700mAh battery for their FreeRunner Feb 01 19:59:39 Kero: the 01 batteries are ~1200 mAh Feb 01 19:59:49 ok. Feb 01 20:05:37 it's still not possible to turn off the display, is it (not talking about backlight)? Feb 01 20:06:45 sure Feb 01 20:06:49 That's in settingsgui Feb 01 20:07:04 And you can just send the right string to the right proc file to do it programatically Feb 01 20:07:15 Oh, wait... Feb 01 20:07:39 To my mind there are three levels of "screen on": lcd off, dim, and full backlight Feb 01 20:07:56 I assume that if the screen is on at all, you can see it... Feb 01 20:08:05 If that's true, what I said earlier is also true Feb 01 20:08:14 If not, then I have no idea. Feb 01 20:08:37 It certainly looks in settingsgui like you have two options: adjust level of backlight, and turn off lcd altogether Feb 01 20:13:32 batteries they are testing right now are 1250 it seems http://wiki.openmoko.org/wiki/Prototypes Feb 01 20:14:17 3.7V Feb 01 20:22:36 br3n: Hmm, that worries me Feb 01 20:22:50 I hoped the FreeRunner would come with those 1700 mAh they were talking about Feb 01 20:23:07 it might :-) Feb 01 20:23:10 1250 is not a standard 01 battery; they're 1200 Feb 01 22:01:11 hi all Feb 01 22:02:10 I have built openmoko and was trying to flash qemu but the following error arise http://paste.ubuntu-nl.org/54389/, could any one help me with this Feb 01 22:03:34 ? Feb 01 22:04:09 madel, looking (but newbie so don't expect much) Feb 01 22:04:50 br3n: ok, thank you Feb 01 22:07:00 It doesn't make any sense to me Feb 01 22:07:12 Have you googled for 'bad block qemu'? Feb 01 22:07:15 :-) Feb 01 22:07:20 madel, i don't know enough to know what is wrong....all i can say is following these instructions exactly worked for me http://wiki.openmoko.org/wiki/Automatic_emulation_in_Ubuntu Feb 01 22:07:34 is that what you followed Feb 01 22:08:45 anyone have experience of serial terminals? Feb 01 22:09:48 ok, let's see if this 2.6.24 kernel boots Feb 01 22:10:43 it does, so the gta01-not-booting bug is really fixed Feb 01 22:14:07 urgh, lockup, before I could manage to get to the cpufreq directory to start testing Feb 01 22:14:48 so how are things going in general? What are we looking at in terms of when the neo will be stable enough for a daily user like me? Feb 01 22:16:06 tushyd, http://wiki.openmoko.org/wiki/Snapshot_review Feb 01 22:17:01 br3n, thanks Feb 01 22:17:15 tushyd, np Feb 01 22:17:57 tushyd, several times a week there will be updates of the latest "daily build" on that page so you can keep watching..... Feb 01 22:18:03 tushyd: that's 2.6.24, the kernel currently being used is 2.6.22.5, which works fine Feb 01 22:18:16 tushyd: I'm hacking on 2.6.24 currently, that's all Feb 01 22:20:39 cesarb, how is cpufreq going? Feb 01 22:22:41 br3n: I'm trying to test it on 2.6.24, but it locks up before I can type the commands to enable it Feb 01 22:23:08 br3n: it's before my code gets active, so I currently have no idea why is that happening Feb 01 22:23:20 br3n: it's a hard lock up, since holding power powers down Feb 01 22:23:22 :-( Feb 01 22:23:23 so cesarb, the 2.6.22.5 is stable? I just want to know if I should buy one... i'm not that into programming but i would consider myself a poweruser Feb 01 22:23:53 tushyd: last I looked at the userspace, it wasn't poweruser-ready yet, but that was more than a month ago Feb 01 22:23:56 tushyd, wait for new freerunner soon to be released Feb 01 22:24:06 cesarb: It looks like the bleeding-edge build has now switched to 2.6.24 - my MokoMakefile is currently compiling "linux-openmoko-2.6.24+svnr4009-r3950-r1" Feb 01 22:24:37 ok, that's cool. I'm willing to wait for it :) Feb 01 22:25:37 tushyd, maybe march maybe april maybe may..... (they say march approx at mo) Feb 01 22:26:12 mmontour: good, is there any success report with gta01 running it yet? Feb 01 22:26:22 is that when gta02 comes out? Feb 01 22:26:42 yes, see topic at irc login Feb 01 22:27:06 mmontour: I built 2.6.24 outside the tree, but other than the cpufreq patches (which should be inert before I enable it on /sys), it should be the same, and it's crashing reliably after about a minute Feb 01 22:27:26 mmontour: I would like to know if it's just me Feb 01 22:27:34 cesarb, at least is reliable :-) Feb 01 22:30:10 * cesarb ponders a bit and guesses it might be the modem not powering up and the console code messing up due to sysfs name changes Feb 01 22:30:25 That serial console code is a mess Feb 01 22:31:39 cesarb: I'm at work right now so I won't be able to test it until this evening, but if it crashes I'll say something here and/or on the mailing list. Feb 01 22:31:51 mmontour: thanks Feb 01 22:32:36 cesarb, only qemu here.... Feb 01 23:32:55 further proving it was the wacky serial code again, I booted without the serial console and it didn't lock up Feb 01 23:38:21 cesarb: some of the neo-specific nodes under /sys changed names in 2.6.24, so you're going to encounter some userspace bugs Feb 01 23:40:16 oh Feb 01 23:40:20 you're aware of that Feb 01 23:40:28 Vegar: it probably was exactly that: it didn't turn on the modem Feb 01 23:40:44 oh well, my code which was working is still working, which is what I wanted to check Feb 01 23:40:55 is the display flicker gone too? Feb 01 23:40:58 but the frambuffer code, which I didn't check before, is completely broken Feb 01 23:41:06 not only the flicker ISN'T gone Feb 01 23:41:31 but also it does a VERY psychadelic (sp?) glitch for a couple of seconds during every transition! Feb 01 23:41:53 the screen fades to yellow and then suddely comes back to life at the wrong speed (as can be noticed by the flicker) Feb 01 23:42:09 psychedelic* Feb 01 23:42:45 Vegar: thanks, I never get that word correct in English (I'm used to "psicodélico", which I always get right... :D ) Feb 01 23:43:11 that's even more complicated than the English word Feb 01 23:44:53 openmoko: 03werner * r4010 10/branches/src/target/kernel/2.6.24.x/scripts/ (README build update): Feb 01 23:44:53 openmoko: Updates "scripts" directory for location under /branches, brought in still Feb 01 23:44:53 openmoko: missing files from /trunk. Feb 01 23:45:47 Vegar: it's actually simpler... no "ch", no "y"... Feb 01 23:45:59 Vegar: it's said the same way it's written... Feb 01 23:46:11 do you pronounce the p too? Feb 01 23:46:28 openmoko: 03werner * r4011 10/ (kernel u-boot): Feb 01 23:46:28 openmoko: - /kernel: symlink to branches/src/target/kernel/2.6.24.x Feb 01 23:46:28 openmoko: - /u-boot: symlink to trunk/src/target/u-boot Feb 01 23:46:34 Vegar: (instead of the incredibly random English pronouciation, where each letter can have several completely distinct sounds) Feb 01 23:46:54 haha, yeah, that's true Feb 01 23:48:27 I already noticed that, while I do not have good English pronouciation (probably due to almost never actually speaking in english), other Brazilians can understand me fine. We must make the exact same mistakes :D Feb 02 00:22:12 openmoko: 03werner * r4012 10/branches/src/target/kernel/2.6.24.x/patches/atheros_2_0_function.patch: (log message trimmed) Feb 02 00:22:12 openmoko: ar6k: Beacon filtering enabled once scan is done Feb 02 00:22:12 openmoko: When we're done with scanning, the host doesn't need to get all the Feb 02 00:22:12 openmoko: beacons anymore. Later on, we could fine tune that to only get our Feb 02 00:22:12 openmoko: BSS beacons (mandatory for roaming). Feb 02 00:22:16 openmoko: But for now, we just filter them all after scanning is done so that Feb 02 00:22:18 openmoko: we don't get dozens of interrupts per seconds because of that. Feb 02 02:50:57 Hello Everybody Feb 02 02:51:32 Hello Everybody Feb 02 02:57:25 Hello Everybody Feb 02 02:57:29 Anybody there? **** ENDING LOGGING AT Sat Feb 02 02:59:57 2008