**** BEGIN LOGGING AT Sat Apr 07 02:59:57 2007 Apr 07 03:04:59 <_ASK> openmoko applications depend on OpenMoko API or they also depend on the underlying OS? Apr 07 03:48:14 Germany had lots of christian holidays. Apr 07 03:48:36 THey had holidays for events on teh christian calendar I'd never heard of. Apr 07 04:19:10 counter Apr 07 04:19:10 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 2 days 03:50:25 (9.160 +-9.2 days) (773,148) Apr 07 04:59:48 openmoko mailing lists are down to - I just got a 6 hour delayed message. Apr 07 07:32:59 good morning Apr 07 07:34:35 morning Apr 07 07:36:02 g'morning. Apr 07 08:44:55 counter Apr 07 08:44:55 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 2 days 01:37:32 (9.068 +-9.1 days) (774,148) Apr 07 09:01:20 re Apr 07 09:09:55 morning Apr 07 09:13:48 morning Apr 07 09:15:46 openmoko: 03laforge * r1682 10/trunk/src/target/kernel/patches/gta01-core.patch: start power-down after three, rather than five seconds (Closes: #418) Apr 07 09:22:18 (script) openmoko-devel: Harald Welte Re: qemu-neo1973 Apr 07 09:26:18 (script) wiki RecentChanges: [[Main_Page/fr]] [[Main_Page]] [[Style_Sheet]] [[Help:Contents]] [[Community_Events]] [[Template:Languages]] [[Main_Page/kr]] [[Hardware/de]] [[Hardware/zh_cn]] [[Documentation]] [[Category:Neo1973_Hardware]] [[FAQ/ru]] [[FAQ/zh_cn]] and other changes Apr 07 09:47:57 counter Apr 07 09:47:57 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, 2 days 01:06:01 (9.046 +-9.0 days) (775,148) Apr 07 09:48:04 looks like everything is back up again. Apr 07 09:59:28 * buz clobbers his tree Apr 07 10:12:01 mickey|zzZZzz: fetching of uboot-gta fails Apr 07 10:24:26 zecke: fetched here ok - svn.openmoko.org was down for a while today Apr 07 10:24:41 but it's up again now Apr 07 10:27:39 rwhitby: http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=OpenMoko somehow there is another try/except missing.... Apr 07 10:34:59 * CM also clobbers and rebuilds :) Apr 07 11:03:16 openmoko: 03mickey * r1683 10/trunk/oe/packages/openmoko-apps/openmoko-simplemediaplayer_svn.bb: oe: openmoko-simplemediaplayer needs libid3, not id3lib (yes, that's a different library) Apr 07 11:03:39 openmoko: 03mickey * r1684 10/trunk/oe/packages/libid3/ (. libid3_1.1.bb): oe: add libid3, a small library for reading id3 tags Apr 07 11:12:45 openmoko: 03mickey * r1685 10/trunk/oe/packages/openmoko-apps/openmoko-simplemediaplayer_svn.bb: oe: revert last change Apr 07 11:15:59 openmoko: 03mickey * r1686 10/trunk/src/target/OM-2007/applications/openmoko-simplemediaplayer/beep/ (Makefile.am bmp.desktop.in): openmoko-simplemediaplayer: canonize executable name Apr 07 11:19:52 openmoko: 03mickey * r1687 10/trunk/src/target/OM-2007/applications/openmoko-simplemediaplayer/beep/ (Makefile.am openmoko-simplemediaplayer.png): openmoko-simplemediaplayer: add .desktop icon Apr 07 11:22:18 (script) wiki RecentChanges: [[Talk:OpenMoko_under_QEMU]] [[WiFi_support_in_OpenMoko]] Apr 07 11:24:49 openmoko: 03mickey * r1688 10/trunk/src/target/OM-2007/applications/openmoko-simplemediaplayer/beep/Makefile.am: openmoko-simplemediaplayer: properly ship png for .desktop file Apr 07 12:51:26 openmoko: 03mickey * r1689 10/trunk/oe/packages/tasks/task-openmoko.bb: oe: add fbgrab into the default image (task-openmoko-debug) Apr 07 12:55:09 mickey, could you trigger a rebuild on the buildhost? Apr 07 12:57:46 Shoragan: I'll launch a full rebuild in my own directory to get your the latest image Apr 07 12:58:03 will take a couple of hours until completion though Apr 07 12:58:23 thanks, it'll be a lot faster than on my thinkpad, tough Apr 07 12:59:05 most likely, ya :) Apr 07 12:59:06 mickey|meeting - Greet Sean from the channel, please. Apr 07 13:05:16 aac playing seems to tabe about 33%cpu at 200mhz Apr 07 13:05:28 tested with mplayer on henryks moko right now Apr 07 13:06:39 Thanks. Apr 07 13:06:47 (if that was in response to my question). Apr 07 13:07:00 So, comparable to ogg, a little less than MP3? Apr 07 13:07:57 nope.. mp3 takes (depending on the player, decoded with libmad) about 20-25% Apr 07 13:08:36 mpg123 (floats need to be emulated) about 30% Apr 07 13:09:19 we need to do testing with libtremor though, libogg is not really optimized at all for !float platforms Apr 07 13:11:08 I thought there was an integer decoder? Apr 07 13:11:19 yep. tremor Apr 07 13:11:22 sorry :) Apr 07 13:11:31 n.p. Apr 07 13:12:37 hi Apr 07 13:12:48 when trying to run qemu Apr 07 13:12:55 i get this error Apr 07 13:13:08 qemu: fatal: Trying to execute code outside RAM or ROM at 0x38000000 Apr 07 13:13:18 any idea? Apr 07 13:13:31 qemu-system-arm -M neo -m 512 -mtdblock openmoko/openmoko-flash.image -kernel openmoko/openmoko-kernel.bin -usb -show-cursor Apr 07 13:13:35 i use this command Apr 07 13:14:03 qemu-system-arm -M neo -m 128 -mtdblock openmoko/openmoko-flash.image -kernel openmoko/openmoko-kernel.bin -usb -show-cursor Apr 07 13:14:15 not 512m, sorry Apr 07 13:14:31 but it does not make any difference Apr 07 13:14:41 arm-softmmu/qemu-system-arm -M neo -m 130 -mtdblock openmoko/openmoko-flash.image -kernel openmoko/uImage-2.6-moko8-r0_0_1637_0-fic-gta01.bin -usb -show-cursor -snapshot is what i use Apr 07 13:16:33 i got the same error Apr 07 13:17:13 hm. which image do you use and what is the system you compile on? Apr 07 13:21:52 i use the same image as you Apr 07 13:22:06 my system is x86 Apr 07 13:22:11 with linux pardus Apr 07 13:24:22 never heard of Apr 07 13:25:33 :) Apr 07 13:25:51 a turkish national distribution Apr 07 13:26:02 i did what it say in the wiki Apr 07 13:26:13 did you use mokomakefile? Apr 07 13:26:27 i only typed make Apr 07 13:27:09 do i need to change anything Apr 07 13:27:27 could the makefile be for arm systems Apr 07 13:27:30 ? Apr 07 13:28:11 http://wiki.openmoko.org/wiki/OpenMoko_under_QEMU Apr 07 13:28:22 i read this page Apr 07 13:30:08 ah.. ok.. so only the emu with a prebuilt moko-image Apr 07 13:30:31 did you configure it with the params as said there? Apr 07 13:30:53 and which version is your gcc Apr 07 13:38:18 mickey|meeting: is g_object_new memsetting the allocated area? Apr 07 13:43:57 :D Apr 07 13:44:02 i did not have lynx Apr 07 13:49:39 uh Apr 07 13:55:25 hi there Apr 07 14:04:53 mickey, where will the images appear? Apr 07 15:03:57 hi Apr 07 15:04:36 i heard the neo 1973 will only have usb 1.1. is an upgrade to usb 2.0 planned for this phone? Apr 07 15:04:37 baze: 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.) Apr 07 15:07:30 one wonders fwhat "this phone" means Apr 07 15:07:35 certainly not the p1 version Apr 07 15:14:57 ok sorry. i meant the p2 one for the mass market Apr 07 15:15:18 (script) wiki RecentChanges: [[Main_Page/kr]] [[UI_Improvements]] Apr 07 15:15:56 but the faq in the wiki almost answered the question :) Apr 07 15:20:59 come to think of it, since the usb interface is coupled into the SOC, and the SOC will be upgraded, wonder if the new one does 2.0... Guess that would've been mentioned though. Apr 07 15:40:20 usb otg would be more important. Apr 07 15:54:00 Elrond: AFAIR one core dev said there would not be USB OTG support... Apr 07 15:55:21 not real otg, but it can do client as well as host Apr 07 15:55:38 host is unpowered, so you'ld need something to poer the client Apr 07 15:55:49 s/poer/power/ Apr 07 15:55:49 roh meant: host is unpowered, so you'ld need something to power the client Apr 07 16:02:56 roh - I know that. But OTG has other aspects than power: OTG can detect on its own, wether it should be client or host. (The cables have trickery for this to work) Apr 07 16:03:14 morning Apr 07 16:03:18 Hi hrw. Apr 07 16:03:24 Hello! Apr 07 16:03:36 neo1973 is not OTG capable anyway Apr 07 16:05:01 maybe version after hardware update will be OTG capable but current (Bv3/Bv4) are not Apr 07 16:07:09 hrw - We were talking about hw-upgrade. Apr 07 16:07:37 ok Apr 07 16:08:00 many people suggest that it will give us s3c2440 or 2443 even Apr 07 16:09:03 Yeah, it looks like that. Did you already take a look at the specs? Apr 07 16:09:50 no Apr 07 16:11:14 and will not check it here due to conection speed limit Apr 07 16:12:13 :) Apr 07 16:17:08 Shoragan: I'll put them under people.openmoko.org/mickey once they're done Apr 07 16:17:11 one or two hours more Apr 07 16:18:03 if you need an image to test i can copy a fresh one on an http Apr 07 16:18:19 mickey|meeting: you are MIA in #opie.de Apr 07 16:25:34 Hi Apr 07 16:25:36 counter Apr 07 16:25:36 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, a day 21:47:11 (8.908 +-8.9 days) (776,148) Apr 07 16:25:50 A week or two. Apr 07 16:25:57 openmoko: 03zecke * r1690 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/src/ (rfcdate.c rfcdate.h): openmoko-rssreader: Add a simple untested RFC822 scanner to be used by the TreeModel Apr 07 16:27:46 or more or less Apr 07 16:28:27 Hi speedy :P Apr 07 16:32:19 counter Apr 07 16:32:20 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, a day 21:43:50 (8.905 +-8.9 days) (777,148) Apr 07 16:40:51 hiya..following the announcement about OE freeze, i did a clobber..but when i do a make update i am getting this Apr 07 16:40:54 svn: Working copy '.' locked Apr 07 16:40:54 svn: run 'svn cleanup' to remove locks (type 'svn help cleanup' for details) Apr 07 16:40:54 make: *** [stamps/patches] Error 1 Apr 07 16:41:24 of course i did run 'svn cleanup .' Apr 07 16:52:08 Esto es lo que se llama un éxito Apr 07 17:07:14 openmoko: 03mickey * r1691 10/trunk/oe/packages/tasks/task-openmoko.bb: oe: ship openmoko-appmanager in task-openmoko-base Apr 07 17:10:52 well never mind...i flushed bitbake/ and all is well Apr 07 17:11:18 (script) wiki RecentChanges: [[Widget_Inheritance_Graph]] Apr 07 17:19:22 * aevin notices the number of subscriptions at planet.openmoko.org has increased from 10 to 11. Apr 07 17:29:43 Shoragan: try the lastest stuff in buildhost.openmoko.org:/tmp/deploy/images Apr 07 17:30:35 mickey|meeting: how goes meeting? Apr 07 17:31:05 thx, mickey :) Apr 07 17:31:08 mickey|meeting: maybe some suggestions to counter value? Apr 07 17:31:13 counter Apr 07 17:31:13 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, a day 21:14:23 (8.885 +-8.9 days) (778,148) Apr 07 17:32:09 so you build on buildhost instead of your own checkout? Apr 07 17:32:23 yes Apr 07 17:32:43 my tree always has local diffs Apr 07 17:34:26 hi Apr 07 17:34:43 CoreDump|home: how goes your neo polishing? Apr 07 17:35:08 hrw: there really isn't anything to polish yet, now is it? ;) Apr 07 17:35:22 (the last rootfs from the buildhost was _horrible_) Apr 07 17:36:06 CoreDump|home: I do not use openmoko distro on neo ;D Apr 07 17:36:22 hrw: heh! What are you using? Angstrom? Apr 07 17:36:25 hrw: lu Apr 07 17:36:31 CoreDump|home: yes Apr 07 17:36:32 hrw: lfs? Apr 07 17:36:58 hrw: interesting. Apr 07 17:37:16 CoreDump|home: it works Apr 07 17:37:23 hrw, howz angstrom behaving on neo ? Apr 07 17:37:56 empty_mind: correctly. it boots, it works Apr 07 17:38:06 hrw: well, the basic OM distro works as well. It's just that you can completely forget the OM applications ontop of matchbox ATM :\ Apr 07 17:38:22 hrw, what did you specify as target nachine ? Apr 07 17:38:32 geee Apr 07 17:38:36 empty_mind: fic-gta01 Apr 07 17:38:41 what target machine for the neo.... Apr 07 17:38:48 that's a hard question Apr 07 17:38:51 hrw, interesting Apr 07 17:38:54 I'll get back to you next week Apr 07 17:39:28 koen, i will wait Apr 07 17:39:47 Apr 07 17:40:21 that means i can run angstron in qemu-neo ? Apr 07 17:40:31 angstrom* Apr 07 17:40:54 hrw: If there were replacements for the OM phone-apps I'd be tempted. But IMO the OM stack should be made to work first before I experiemt w/ different distros below OM Apr 07 17:41:24 * empty_mind wonders if angstrom is working on neo then why are we distro :-D Apr 07 17:42:02 empty_mind: to have more strict control over images content? Apr 07 17:43:13 hrw, hmm, cant we do tha using ipkg and angstrom ? Apr 07 17:43:25 openmoko: 03mickey * r1692 10/trunk/src/target/OM-2007/applications/openmoko-calculator/data/Makefile.am: openmoko-calculator: ship .desktop icon in correct directory Apr 07 17:44:30 CoreDump|home: openmoko should hire a dozen more experienced FOSS/gtk coders first Apr 07 17:45:17 instead of the shanghai lab whose credo seems to be "why patch if you can fork" Apr 07 17:45:19 Hmm, the wiki can't do redirects? Apr 07 17:45:29 that is probably true Apr 07 17:45:29 koen, or they can use python-openmoko when its ready :-p Apr 07 17:45:59 python on a 200mhz arm cpu Apr 07 17:46:01 right.. Apr 07 17:46:21 koen, i am working on that btw Apr 07 17:47:08 koen, we need to do loads of optimization and thoughtfully use cross compilation flags Apr 07 17:47:24 OE already has that flags Apr 07 17:47:34 * koen is "not impressed" Apr 07 17:48:00 koen, yup but having python bindings for openmoko libs is another thing Apr 07 17:48:05 openmoko: 03mickey * r1693 10/trunk/oe/packages/openmoko-pim/ (files/openmoko-dates.desktop openmoko-dates_svn.bb): oe: openmokify dates .desktop Apr 07 17:48:25 *alarm bells* Apr 07 17:48:34 koen, and we need to reduce the footprint too Apr 07 17:48:36 "openmokify .desktop" Apr 07 17:48:44 *alarm bells* Apr 07 17:49:26 * koen hopes openmoko didn't invent its own .desktop format Apr 07 17:49:40 * CoreDump|home doubts it Apr 07 17:49:53 heh Apr 07 17:49:57 you guys have crazy ideas Apr 07 17:50:03 * empty_mind hopes so too Apr 07 17:50:23 mickey|meeting: you're still a (asian) company Apr 07 17:50:29 mickey|meeting, its less crazy than running asterix on openmoko Apr 07 17:50:39 koen: I bet it is e17's desktop format Apr 07 17:50:39 companies like that have crazy notions Apr 07 17:50:55 zecke: which uses the fdo spec nowadays Apr 07 17:51:02 they dumped .eapps Apr 07 17:51:15 thank god Apr 07 17:51:21 .eapps was a PITA Apr 07 17:51:26 indeed Apr 07 17:52:40 mickey, this image is a lot better than the last one from buildhost.o.o Apr 07 17:52:58 nod Apr 07 17:53:07 I've been working on a demo for Sean the whole night Apr 07 17:53:09 TIRED Apr 07 17:53:48 :) Apr 07 17:54:15 * SpeedEvil passes mickey|meeting 6 hours of sleep. Apr 07 17:55:35 openmoko: 03mickey * r1694 10/trunk/oe/packages/openmoko-pim/files/ (Dates.png openmoko-dates.png): oe: catch up with icon rename Apr 07 17:56:05 rwhitby are you there? Apr 07 17:56:40 roh: he probably sleep now Apr 07 17:56:48 0400 in au Apr 07 17:56:52 Afternoon roh :) Apr 07 17:57:00 03:25 at rwhitby Apr 07 17:58:05 openmoko: 03mickey * r1695 10/trunk/oe/packages/openmoko-base/files/session: openmoko-session: launch all applets Apr 07 18:01:20 hey dho Apr 07 18:02:42 cu Apr 07 18:03:03 * CoreDump|home shall install the new rootfs tomorrow Apr 07 18:04:08 btw, I really have to tell you as feedback: This debug-board cable thingy _really_ honestly to $DEITY sucks :\ Apr 07 18:04:37 CoreDump|home we know *sigh* Apr 07 18:04:44 wiki: How clever was it to use category pages for real content? Apr 07 18:04:49 i fought with it also about 6 hours the last 2 days Apr 07 18:04:53 CoreDump|home: it only took me 5 hours to get it to connect properly Apr 07 18:04:58 It looks and feels so fragile I don't really dare using (or removing) it Apr 07 18:05:23 koen: do that like 5 to 10 times and you can throw it away :\ Apr 07 18:05:28 it's rated at 20 insertions, so don't brick you neo to often... Apr 07 18:05:33 counter Apr 07 18:05:33 http://wiki.openmoko.org/wiki/Neo1973 P1: a week, a day 20:57:13 (8.873 +-8.9 days) (779,148) Apr 07 18:05:50 monotone can be so Slow to update the tree. Apr 07 18:06:00 /can be// ? Apr 07 18:06:02 _is_ Apr 07 18:06:07 :) Apr 07 18:06:18 all that silly rsa Apr 07 18:06:25 im at 1,676/4,445 "certs in" and its been running for 30 minutes. Apr 07 18:06:56 and then it'll tell you that you need to clobber it and start over Apr 07 18:07:03 lol Apr 07 18:07:15 first time it takes about 6-7 hours on a fast (core duo 2ghz, 1g ram) computer with about 4mbit sdsl ip Apr 07 18:07:16 have you seen that message yet? Apr 07 18:07:25 not yet. Apr 07 18:07:39 i was encouraged by the latest planet posting about 'make qemu' Apr 07 18:07:56 so i dusted off my last attempt to build openmoko Apr 07 18:08:28 and im waiting through make update-makefile; make update Apr 07 18:09:11 so maybe i can atleast pretend to own a neo1973 :) Apr 07 18:09:37 tihi. Apr 07 18:14:15 hi all Apr 07 18:14:29 i ask again, sorry i forget... what program created this: http://wiki.openmoko.org/images/2/28/Openmokoframework.png ? Apr 07 18:14:46 visio? Apr 07 18:14:49 or dia? Apr 07 18:14:56 maybe even omnigraffle Apr 07 18:15:09 windows program? hard to think.. Apr 07 18:15:10 duh! it was paint :-P Apr 07 18:15:14 :D Apr 07 18:15:48 i would think omnigraffle ;) Apr 07 18:16:12 they to use an ancient eds-dbus version Apr 07 18:16:16 2060126 Apr 07 18:16:20 i want something like this on linux Apr 07 18:16:28 I wonder if the romans already had tinymail Apr 07 18:16:29 i see, is there a dia named program, i try out:) Apr 07 18:22:38 metamorgozis: www.gliffy.com ? Apr 07 18:23:04 once you created it, you can export as SVG or PNG. The SVG can still be editted with inkscape Apr 07 18:23:13 yes i know, cool Apr 07 18:23:26 actually i see dia, its cool except it can't do that lines that on that image... Apr 07 18:23:42 the -------/\----->'s Apr 07 18:23:43 :D Apr 07 18:24:05 It does that for you automatically when lines intersect Apr 07 18:24:37 i havenbt looked at dia for years. it sucked when i looked at it. i think inkscape is better Apr 07 18:25:06 inkscape is cool Apr 07 18:25:14 but not for this Apr 07 18:25:15 :) Apr 07 18:26:00 i can appreciate being able to build the platform, but is there a pre-built image for linux that i can download and run? Apr 07 18:26:56 MacNorth: there it is: "mtn: fatal: std::logic_error: roster.cc:1,695: invariant 'I(new_roster == from_right_r)' violated. this is almost certainly a bug in monotone." Apr 07 18:28:00 * don-o starts make update again Apr 07 18:28:24 maybe an ubuntu user can tar up their /home/moko? Apr 07 18:29:29 didn't someone here create a vmware image? Apr 07 18:31:21 don-o: i have feisty installed and just built it Apr 07 18:31:25 but it's huge Apr 07 18:31:36 you could however use the prebuilt images in qemu-arm Apr 07 18:32:13 mokomakefile has targets to do that Apr 07 18:33:56 i'll try. out of curiosity how big is the image you built Apr 07 18:34:20 the image itself is not very big Apr 07 18:34:23 the directory however... Apr 07 18:34:33 er yeah ,the whole directory. Apr 07 18:34:48 don-o: still running du on it ;) Apr 07 18:34:54 ha Apr 07 18:34:57 it's huge by any means Apr 07 18:35:21 8,9GB Apr 07 18:35:34 downloading something huge would still beat wrestling monotone and my slow build computer. Apr 07 18:35:35 don-o a complete built moko tree can take 10-16 gbyte, depending how much revisions youve build Apr 07 18:35:42 yeah thats pretty big. Apr 07 18:36:07 i can tar up my qemu with built flash if you want Apr 07 18:36:40 roh: sounds good Apr 07 18:37:10 buz: IMHERIT += "rm_work" Apr 07 18:37:25 keeps it down to 300MB Apr 07 18:37:42 you sure it shouldnt be INHERIT? Apr 07 18:37:51 yeah Apr 07 18:37:56 turing test ;) Apr 07 18:37:58 weird spelling Apr 07 18:38:33 doesnt really matter Apr 07 18:38:39 teh disk currently is big enough Apr 07 18:39:47 in the pdf version Apr 07 18:39:50 i found it's (OmniGraffle Professional 4.1.2)/ModDate Apr 07 18:41:41 it's a non free mac prog. Apr 07 18:41:44 brah! Apr 07 18:41:55 don-o will take a while to upload it.. i'm here at eh07 which is 250 geeks sharing a 2mbit sdsl line Apr 07 18:42:09 openmoko: 03mickey * r1696 10/trunk/src/target/OM-2007/applications/openmoko-appmanager/is-processing.patch: openmoko-appmanager: remove is-processing patch from here, it's now in oe/packages/ipkg Apr 07 18:42:12 roh: k. thx. Apr 07 18:42:51 omigraffle is the shizzle Apr 07 18:42:57 s/omi/omni Apr 07 18:57:41 openmoko: 03zecke * r1697 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/src/ (rfcdate.c rfcdate.h): openmoko-rssreader: Minor change add void as parameter (C++->C is hard) Apr 07 18:59:41 openmoko: 03zecke * r1698 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/ (5 files in 2 dirs): Apr 07 18:59:41 openmoko: openmoko-rssreader: Start adding unit-tests for the non gui bits Apr 07 18:59:41 openmoko: Start adding unit tests for the RSS RFC Date class. This mostly Apr 07 18:59:41 openmoko: sets up autofoo to easily add new tests. Apr 07 19:03:45 don-o http://yamato.hyte.de/moko/qemu-neo1973.tar.bz2 59731768 bytes Apr 07 19:03:50 connecting neo to debug board is easy Apr 07 19:04:25 connector suxx anyway Apr 07 19:05:04 anyways.. i'm doff to the bar for a while ;) Apr 07 19:05:06 -d Apr 07 19:05:26 * hrw remember 90GB builds... Apr 07 19:05:47 ^ roh that image is runnable with quemu? Apr 07 19:05:56 i don't follow the conversation Apr 07 19:06:18 (script) wiki RecentChanges: [[Hardware]] [[Talk:OpenMoko_under_QEMU]] Apr 07 19:07:23 MetaMorfoziS: recent openmoko-qemu can boot neo1973 images Apr 07 19:07:44 * koen wonders if openmoko-qemu has the nptl patch Apr 07 19:08:16 check? Apr 07 19:08:34 openmoko: 03zecke * r1699 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/ (. data/ po/ src/ tests/): openmoko-rssreader: Ignore autogenerated files Apr 07 19:08:45 hrw: how? Apr 07 19:09:06 hrw: instead of patches it's another fork in svn.... Apr 07 19:09:32 I see Apr 07 19:09:34 fsck.. andrew keep whole qemu patched dir in svn without patchsets... Apr 07 19:09:40 right Apr 07 19:09:49 just like every other fork in openmoko svn Apr 07 19:09:52 hehe Apr 07 19:10:27 eventually there will be patches Apr 07 19:10:36 and I fear OM developers are too busy with sorting out that mess and push patches upstream the next 6 months Apr 07 19:11:00 koen: andrew is OH guy Apr 07 19:11:01 s/ with sorting out that mess/to sort out that mess/ Apr 07 19:11:01 koen meant: and I fear OM developers are too busyto sort out that mess and push patches upstream the next 6 months Apr 07 19:11:15 hrw: I know Apr 07 19:11:22 and he is pushing his qemu things upstream Apr 07 19:11:32 cool Apr 07 19:11:46 he wrote spitz emulation Apr 07 19:11:47 so there are patches we can add to qemu-native in OE? Apr 07 19:11:55 hrw: I also know that Apr 07 19:12:13 koen: wait a bit and I will push such thing Apr 07 19:12:20 or a bit longer then a bit ;D Apr 07 19:12:49 as I do not know when I will work on updating qemu (and does it will be me or not) Apr 07 19:14:31 hmm I need a faster machine Apr 07 19:21:19 openmoko: 03mickey * r1700 10/trunk/src/target/OM-2007/applications/openmoko-appmanager/src/ (application-menu.c appmanager-window.c): openmoko-appmanager: set UI-visible application name, fix some typos Apr 07 19:34:17 openmoko: 03zecke * r1701 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/ (configure.ac po/LINGUAS po/POTFILES.in): Apr 07 19:34:17 openmoko: openmoko-rssreader: Make i18n work(?!) Apr 07 19:34:17 openmoko: Require intltool and provide an empty LINGUAS file Apr 07 19:35:08 openmoko: 03zecke * r1702 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/po/: openmoko-rssreader: ignore autogenerated i18n files Apr 07 19:42:29 cjb-: hah, another cjb Apr 07 19:43:09 zecke: any html widget yet in rssreader? Apr 07 19:45:15 koen: I will wait until the 11th before I decide what to do Apr 07 19:45:28 that's after your exams? Apr 07 19:45:43 cu Apr 07 19:45:45 koen: 11th is the GSoC deadline Apr 07 19:45:54 ah, right Apr 07 19:46:19 koen: exams? I don't take the written ones, I'm 'special' and still need to chase the profs Apr 07 19:46:22 speaking of deadlines Apr 07 19:46:30 * koen registers for some new courses Apr 07 19:46:31 koen: when do you start packing? Apr 07 19:46:43 zecke: next week Apr 07 19:47:45 although I'm at my parents now for easter Apr 07 19:48:28 (actually because they live close to the wildlife reserve we're going to visit tomorrow) Apr 07 19:49:53 koen: I think I was there once. I assume there are not too many reserves in the nl Apr 07 19:50:15 you were probably close Apr 07 19:50:27 this is next to the big reserve :) Apr 07 19:50:49 the small reserve is the former hunting ground for the royal family Apr 07 19:51:45 zecke: was it 'de veluwe' in the middle of .nl? Apr 07 19:53:10 koen: http://www.openembedded.org/~zecke/safari.jpg I don't know where it was, but they sold me this hat Apr 07 19:53:37 aha! Apr 07 19:53:40 :) Apr 07 19:53:46 "safaripark beekse bergen" Apr 07 19:53:52 I should visit that again Apr 07 19:54:06 probably :) Apr 07 19:55:56 heh Apr 07 19:55:57 http://bsd.slashdot.org/comments.pl?sid=229865&cid=18647919 Apr 07 19:58:18 koen: bergen? Apr 07 19:58:20 cjb: sometimes I wish that unix account names were allocated by a central global authority. :) Apr 07 19:59:04 zecke: http://www.safaripark.nl/home1.aspx Apr 07 19:59:38 koen: I'm already watching the animals Apr 07 20:09:09 zecke: http://sayamindu.randomink.org/openmoko_browser.png Apr 07 20:09:53 koen: using libgtkhtml? Apr 07 20:10:34 koen: GtkWebCore Apr 07 20:11:32 gtk-webcore Apr 07 20:11:51 koen: I know. I see /usr/include/osb and BridgeImpl::windowObjectCleared Apr 07 20:14:32 hmm,will you be able to browse websites the other way too ? (eh not like this | like this - ) Apr 07 20:17:59 zecke: you can see it's webcore because of the mandatory scrollbars Apr 07 20:18:54 hehe Apr 07 20:31:44 openmoko doesn't use cairo, does it? Apr 07 20:32:41 baze: it's availabl Apr 07 20:32:42 e Apr 07 20:33:37 but it's not used as default or what does that mean? Apr 07 20:37:58 baze: Last I heard, openmoko was using a pre-cairo version of GTK+ (2.6). Apr 07 20:38:48 yeah, that's my information aswell Apr 07 20:39:02 is it because of the performance or why? Apr 07 20:39:21 cairo could be used to create nice ui, imho, that's why i ask Apr 07 20:43:54 baze: Yes, cairo could be quite helpful. And it may have been for performance reasons originally, but I think with cairo 1.4.x those reasons shouldn't exist. Apr 07 20:44:14 baze: So it's mostly just a matter of people trying it out now and demonstrating it to work well I think. Apr 07 20:45:13 cworth: gtk 2.10 is slow in non-cairo areas as well Apr 07 20:45:27 cworth: altough I'm running gtk 2.10.10 in my neo all the time Apr 07 20:45:45 cworth: it's quite useable with the surface-cache patch Apr 07 20:45:55 * koen prays OM will stay singlethreaded Apr 07 20:46:35 koen: Yeah, there could be other non-cairo-related performance regressions in GTK+ after 2.6. Apr 07 20:46:35 Why - what about multicore CPU phones? Apr 07 20:46:41 koen: thats the plan according what harald told me. use statemachines Apr 07 20:47:08 koen: I did always wonder if cairo was acting as the scapegoat for all GTK+ performance problems. Hopefully that's gone now. :-) Apr 07 20:47:22 roh: the surface-cache patch in OE speeds up cairo *a lot*, it just isn't safe for multithreaded apps using cairo Apr 07 20:47:49 cworth: right now openmoko is getting hit by floatingpoint image scaling in gtk Apr 07 20:48:15 koen: if it's worth a try then we will try it out shurely. Apr 07 20:48:42 roh: it's applied by default Apr 07 20:49:08 oh Apr 07 20:50:54 ?! Apr 07 20:51:06 oops.. Apr 07 20:56:47 koen: hmm the rssreader uses threads :) Apr 07 20:57:32 zecke: and accessing the same cairo-context in the threads? Apr 07 20:57:51 http://lists.freedesktop.org/archives/cairo/2007-April/010308.html Apr 07 20:58:02 koen: not directly, but I'm sure GtKtreeView is using it indirectly Apr 07 21:32:00 MokoMakefile is now correctly building the exact same image as the core development team. Apr 07 21:46:34 rwhitby: kudos, thanks! Apr 07 21:55:33 rwhitby: i think package sjf2410-linux-native-20060807-r1_0 is broken for INHERIT += "rm_work" Apr 07 21:58:30 openmoko: 03zecke * r1703 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/ (src/rfcdate.c tests/date_test.c): openmoko-rssreader: Add more tests and besides the memory leaks of Glib we are doing fine Apr 07 21:59:04 http://pastebin.ca/428881 : do_rm_work happens before do_deploy Apr 07 22:03:43 MacNorth: thank you :) Apr 07 22:04:01 np :) Apr 07 22:05:08 openmoko: 03zecke * r1704 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/: openmoko-rssreader: Ignore more i18n autogenerated files Apr 07 22:44:31 zecke: Apr 07 22:44:33 | aclocal:configure.ac:37: warning: macro `AM_PATH_CHECK' not found in library Apr 07 22:44:33 | autoreconf: running: /home/moko/build/tmp/staging/i686-linux/bin/autoconf --include=/home/moko/build/tmp/staging/arm-linux/share/aclocal-1.9 --include=/home/moko/build/tmp/staging/arm-linux/share/aclocal --force --warnings=cross Apr 07 22:44:33 | configure.ac:37: error: possibly undefined macro: AM_PATH_CHECK Apr 07 22:44:36 (RHEL4) Apr 07 22:44:59 automake (GNU automake) 1.9.2 Apr 07 22:45:10 rwhitby: yeah stupid me :) Apr 07 22:46:01 rwhitby: luckily dropbear is not fetchable atm Apr 07 22:46:58 hmm I wonder why bug reporting is not working ATM Apr 07 22:47:44 ah because the bug was already filed :) Apr 07 22:48:40 rwhitby: good morning and can you do me a favor? Apr 07 22:52:35 zecke: sure, but need to do an easter egg hunt with the kids first. Apr 07 22:52:56 sure Apr 07 22:56:53 rwhitby: http://rafb.net/p/x49VGn55.html check if that magically fixes the issue Apr 07 22:58:18 (script) wiki RecentChanges: [[Buying_Interest_List]] Apr 07 23:13:05 openmoko: 03mickey * r1705 10/trunk/src/target/OM-2007/panel-plugins/openmoko-panel-mainmenu/ (5 files in 2 dirs): openmoko-panel-mainmenu: add enabling code for action button events Apr 07 23:14:09 openmoko: 03mickey * r1706 10/trunk/src/target/OM-2007/devel/qmake/openmoko-include.pro: qmake: define some more vars Apr 07 23:14:30 openmoko: 03mickey * r1707 10/trunk/src/target/OM-2007/openmoko-libs/TODO: openmoko-libs: remove completed stuff from TODO Apr 07 23:14:50 openmoko: 03mickey * r1708 10/trunk/src/target/OM-2007/devel/scripts/launch-xoo: devel/scripts: launch more applets Apr 07 23:15:52 g'night Apr 07 23:15:57 mickey|zzZZzz: night Apr 07 23:41:53 roh: thanks for the image. Apr 07 23:42:08 roh: whats the command line to start qemu using that image? Apr 07 23:43:22 oeh.. dunno mompls Apr 07 23:44:48 uuh.. i tested that with mokomakefile Apr 07 23:45:17 roh: the end of flash.sh has some stuff about how to start qemu Apr 07 23:46:28 don-o i just tar-ed /home/roh/svn/openmoko/openmoko/trunk/src/host/qemu-neo1973 Apr 07 23:46:56 arm-softmmu/qemu-system-arm \ Apr 07 23:46:56 -M neo -m 130 -usb -show-cursor -snapshot \ Apr 07 23:46:56 -mtdblock openmoko/openmoko-flash.image \ Apr 07 23:46:58 -kernel openmoko/openmoko-kernel.bin Apr 07 23:47:14 thats what mokomakefile make run-qemu does Apr 07 23:47:44 there are propably some symlinks in place for that Apr 07 23:48:41 k. im thinking qemu was also a special build. Apr 07 23:48:59 i built the qemu on a slightly updated ubuntu dapper Apr 07 23:48:59 the machine specification, -M neo, says neo is not in the supported machine list Apr 07 23:49:14 don-o with your or my build? Apr 07 23:49:27 dpkg install of qemu Apr 07 23:49:33 lol Apr 07 23:49:36 lol Apr 07 23:49:42 roh oh i see qemu is also in your tarball :) Apr 07 23:50:00 don-o you need to start the qemu in the tarball. it is patched with wthe source also included Apr 07 23:50:00 zecke: I tried putting that file in the workdir/m4 subdir, then run bitbake -c configure, but that didn't help. Apr 07 23:50:15 it nearly completely emulates the real hardware of the moko Apr 07 23:50:24 hi rwhitby Apr 07 23:50:29 hi Apr 07 23:50:30 rwhitby: that is okay (it was a placebo) Apr 07 23:51:05 rwhitby: I try to remember the AC_ option to specify the m4 include dir Apr 07 23:51:12 rwhitby mokomakefile does not build (or only copy, not 100% shure due to clean) dyngen into the right dir Apr 07 23:51:26 roh: it does for me Apr 07 23:51:29 rwhitby so make qemu fails there Apr 07 23:51:40 roh: full pastebin please, cause it works here Apr 07 23:51:57 (on Debian Etch and RHEL4) Apr 07 23:51:59 roh its booting! Apr 07 23:51:59 hay evry1 Apr 07 23:52:18 will do as soon as i can regenerate the situation. i just copied it to the path where it expected it and it worked Apr 07 23:52:30 don-o horay! Apr 07 23:52:33 just wondering when the device will be shipping to the UK, cant w8 to get me hands on one Apr 07 23:52:51 roh: make sure that the trunk/src/host/qemu-1973 is clean (i.e. you didn't configure or build in there) - mokomakefile builds under build/qemu Apr 07 23:52:51 don-o space and return are swaped, contrary to what the docs say (at least for me) Apr 07 23:52:55 It is delayed by extra days every time you use text speak on IRC. Apr 07 23:53:14 shumz: A week or two is the current best guess. Apr 07 23:53:24 Maybe three. Apr 07 23:53:27 roh: latest build docs are correct as far as space and return goes Apr 07 23:53:28 Hopefully not four. Apr 07 23:53:29 roh: i have to hit space to execute "boot". Apr 07 23:53:30 rwhitby ah.. i built it before manually (as described in the wiki at that time) before there was support for it in mokomakefile Apr 07 23:53:33 rwhitby: place ACLOCAL_FLAGS="-Im4" in the autogen.sh Apr 07 23:53:36 lol speed devil Apr 07 23:53:48 don-o yes. should be return says the docs Apr 07 23:53:54 i seroiusly feel like runnig down there and grabbing one Apr 07 23:53:55 but its swaped Apr 07 23:53:59 roh: yeah thta felt unintutitve Apr 07 23:54:12 roh: remove trunk/src/host/qemu-neo1973, svn up, rebuild with mokomakefile Apr 07 23:54:14 speedDevil ill order a few Apr 07 23:54:32 rwhitby: help me, is RHEL4 more outdated than debian stable? Apr 07 23:54:33 * don-o waits for /dev to be fully populated :) Apr 07 23:54:47 roh: the latest flash.sh script prints the correct mappings for space and return Apr 07 23:54:52 rwhitby will do as soon as sober again.Ã.. i'm sitting in the bar at a hacker-meet-and-greet in hamburg with a drink in my left hand Apr 07 23:55:14 SpeedDevil any way i can get notofied on its release? Apr 07 23:55:14 shumz: 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.) Apr 07 23:55:26 rwhitby ;) fixed in the docs instead the source Apr 07 23:55:32 shumz: subcribe to the announce list Apr 07 23:56:07 cheers, i think this device is kinda cheap with all its capabilities nice one dudes Apr 07 23:57:07 zecke: dunno about relative outdatedness between RHEL4 and Debian Sarge. Apr 07 23:57:27 zecke: but it failed the same way on Debian Etch. Apr 07 23:57:55 zecke: sorry, I'm out of time to debug this now - Easter day stuff Apr 07 23:58:00 rwhitby: right, the dependency on check should be optional Apr 07 23:58:12 rwhitby: don't worry, I will remove check on my machine and check it Apr 07 23:58:59 * don-o crashes and burns with a segfault Apr 07 23:59:43 btw: a lot of openwrt guys would like to buy the phone asap and are wanting to build support in their buildsystem .. so eventually we get a second buildsys for free testing our patches Apr 08 00:00:41 don-o qemu crashed? which ubuntu do you use? Apr 08 00:01:18 i want to play with one fo these now :( Apr 08 00:01:28 roh: the startup got as far as some gnome text messages, never switched to X as far as i can tell. im using feisty fawn Apr 08 00:01:29 shumz come over;) Apr 08 00:01:39 roh where are you Apr 08 00:01:57 you lucky "censored" Apr 08 00:01:57 don-o uh.. i see. i'm at dapper with upgrades. feisty as soon as released Apr 08 00:02:24 roh: k. feisty will probably come out at the same time as the neo. Apr 08 00:02:32 shumz berlin usually, atm i'm in Hamburg, henryk is also here with me and also has a gta01bv02 Apr 08 00:02:44 don-o P1 you mean Apr 08 00:03:15 roh: i mean to say 'in a couple weeks' Apr 08 00:03:32 i want one now, is there no way i can get it Apr 08 00:03:44 apart from build it myself Apr 08 00:03:46 shumz: see /topic Apr 08 00:03:50 shumz try the emulator (qemu) Apr 08 00:04:06 its actually faster than the real thing on most actions Apr 08 00:04:21 just the crazy high-res lcd isn't there Apr 08 00:04:50 roh: But the screen emulation in qemu is emulating the display driver on the board, right? Apr 08 00:04:50 im lovin the whole damn idea, it should put a massive dent in the market Apr 08 00:05:01 the qemu window on my 1400x1050 14" screen is >2times as big in X and Y direction than the real one Apr 08 00:05:29 dho it emulates the real hardware, so you get a vertical window showing the same than then screen Apr 08 00:05:36 just be careful alot of clones might be released Apr 08 00:05:55 dho: we boot a unmodified moko kernel and filesytem in the qemu Apr 08 00:05:57 they might steal the code! Apr 08 00:06:30 greghunt: if someone does it better they can if they take care of the gpl of the drivers Apr 08 00:06:45 just go to asia, ull see ur device with a different name for a quarter of the prices Apr 08 00:06:56 shumz fic IS from asia Apr 08 00:07:21 i knew that "as if" lol Apr 08 00:07:26 roh: Right, ok. I was just making sure for the port since I guess we probably wont' have the phones before we start Apr 08 00:07:31 1/w 1 Apr 08 00:08:37 looks like ill have to go play wid da ps3 until i get me hands on this Apr 08 00:10:21 goodnight evry1 cant w8 til this damn thing hits me Apr 08 00:10:31 I wonder how much money I `lose' playing Guitar Hero instead of coding. Apr 08 00:10:44 lol Apr 08 00:11:03 im still learning coding Apr 08 00:11:10 doing a computer science degree Apr 08 00:11:13 * SrRaven can do hello word with java,thats it Apr 08 00:11:21 just cant get me head around java Apr 08 00:12:39 tata, cheerio, toodle doo Apr 08 00:15:44 * dho can't stand people who shorten ``you'' :( Apr 08 00:16:26 * roh reached the tripout of his t60 the 4rth time in a weekend Apr 08 00:22:29 MacNorth: ah, yeah, will look into sjf2410-linux later - will start an rm_work build now Apr 08 00:22:40 openmoko: 03zecke * r1709 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/src/rfcdate.c: Apr 08 00:22:40 openmoko: openmoko-rssreader: There is still something wrong with formating dates Apr 08 00:22:40 openmoko: Somehow I get funny dates such as 29th of February or 31st of Apr 08 00:22:40 openmoko: April. I don't know where they come from, this somehow looks like Apr 08 00:22:40 openmoko: a off by one in month name calculation. This might come from mrss Apr 08 00:22:43 openmoko: and the atom->RSS date conversion. Apr 08 00:25:25 openmoko: 03zecke * r1710 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/src/ (application-data.h callbacks.c main.c): (log message trimmed) Apr 08 00:25:25 openmoko: openmoko-rssreader: Replace the string representation of the date with RSSRFCDate Apr 08 00:25:25 openmoko: Using a GObject for the date allows us to decouple representation to Apr 08 00:25:25 openmoko: the user (locale aware, shortened to Today) and sorting. Apr 08 00:25:26 openmoko: Use the sort_model for the TreeView and set a dedicated data_func and Apr 08 00:25:28 openmoko: sort func for the date column. Apr 08 00:25:30 openmoko: dates are sorted by zsort, and even z doesn't understand the logic Apr 08 00:26:51 * rwhitby starts builds from scratch on Debian Etch and CentOS 4 with rm_work enabled ... Apr 08 00:41:08 openmoko: 03zecke * r1711 10/trunk/src/target/OM-2007/applications/openmoko-rssreader/ (autogen.sh m4/ m4/check.m4): Apr 08 00:41:08 openmoko: openmoko-rssreader: include check.m4 to make autoreconf work on systems without installed check Apr 08 00:41:08 openmoko: Check is used for the unit tests and is an optional depedency, Apr 08 00:41:08 openmoko: include the check.m4 to make sure autoreconf will not fail. Apr 08 00:41:32 rwhitby: happy easter, your build should not stop working at the rssreader Apr 08 00:48:01 zecke: thx. happy easter to you too Apr 08 01:17:23 nite Apr 08 01:46:19 rwhitby: i don't know if it's a mmf 'bug' anyway. the bb recipe is most likely out of order Apr 08 01:47:00 yeah, you will just need to change the ordering of do_deploy Apr 08 01:47:05 it's a common error in the recipes Apr 08 01:47:22 look at openocd or dfu-util for an example of the correct ordering Apr 08 01:48:12 ok, thanks. enjoy your afternoon, get out in the sun! Apr 08 01:57:26 koen|away: i agree with your sentiment -- branches are not preferable to patches Apr 08 01:57:30 le sigh: http://pastebin.ca/429145 Apr 08 01:58:19 Dept. of Redudancy Department -- 2 dfu-util recipes in triplicate Apr 08 02:15:25 openmoko/trunk/oe/packages/dfu-util/dfu-util-native_svn.bb has: addtask deploy before do_package after do_install Apr 08 02:15:40 openembedded/packages/s3c2410-utils/sjf2410-linux-native_20060807.bb has: addtask deploy before do_build after do_compile Apr 08 02:15:52 could use some tweaking ;) Apr 08 02:55:01 MacNorth: that's one disadvantage of the frozen OE version - that problem with sjf2410-linux-native has been fixed in OE head. Apr 08 02:55:25 meh, sooner or later it'll get caught up :) Apr 08 02:55:30 the only expense is hd space Apr 08 02:56:02 even with do_rm_work, du -sh still reports 2.3G Apr 08 02:56:16 I may have to set up MokoMakefile to be able to patch the frozen OE checkout ... **** ENDING LOGGING AT Sun Apr 08 02:59:57 2007