**** BEGIN LOGGING AT Fri Aug 25 02:59:56 2006 Aug 25 06:48:15 03hrw 07org.oe.dev * rfdb9a519... 10/ (3 files in 3 dirs): feed-browser: move section list to end of page source, fixed 2 column layout CSS Aug 25 06:48:20 03hrw 07org.oe.dev * r043ace40... 10/ (1 contrib/feed-browser/index.php): feed-browser: show feed name in download section, sort by feed/arch Aug 25 06:48:26 03hrw 07org.oe.dev * rcb284cc9... 10/ (1 contrib/feed-browser/index.php): feed-browser: show links to all depends/provides/conflicts, show packages not only by Name: but also by Provides: Aug 25 06:58:55 hm this is weird Aug 25 06:59:39 bootstrap-image seems to do want to build a rootfs without all the dependencies like dropbear having been fulfilled, and thus is failing because it can't find all the stuff that should go into the rootfs Aug 25 07:04:47 bonjour Aug 25 07:09:30 hm, there are dependencies like busybox and basefiles in task-bootstrap, but it just doesn't seem to build those Aug 25 07:11:02 03hrw 07org.oe.dev * r1010cdd5... 10/ (1 contrib/feed-browser/update.php): feed-browser/updater: removed warnings, simplified, reformatted Aug 25 07:30:34 giel: I had a similar problem recently... Aug 25 07:30:49 I fixed it by running `bitbake package-index` Aug 25 07:34:01 why is it that I can add a virtual package (virtual/libxine) to DEPENDS, but not to RDEPENDS? Aug 25 07:34:45 from the question alone, its obvious you dont understand the difference between DEPENDS and RDEPENDS at all Aug 25 07:35:08 kergoth: thats not true at all Aug 25 07:35:41 I just don't quite understand how virtual providors work in the context of DEPENDS and RDEPENDS Aug 25 07:36:15 a "virtual provider" isnt special. its just another package, which may or may not provide a binary package, and which tends to not include any content of its own Aug 25 07:36:38 right Aug 25 07:37:49 so, when I say RDEPENDS="libxine" it knows to use the one specifed by PREFERED_PROVIDER_virtual/libxine Aug 25 07:38:39 it will determine which packages have 'libxine' in their PACKAGES, and select the most appropriate one of those based on your preferences, as i understand it. Aug 25 07:38:58 right... Aug 25 07:39:04 but I'm still unclear as to why you say DEPENDS="virtual/package", but not RDEPENDS="virtual/package" Aug 25 07:39:16 then you still dont get what a runtime dependency is Aug 25 07:39:30 kergoth: as said before... that's not true!! Aug 25 07:39:33 its the emitted binary package name, not a bitbake provider Aug 25 07:39:47 apparently it is, or else you wouldnt be asking the question. Aug 25 07:39:47 ok... fine. thats all you had to say! Aug 25 07:40:17 kergoth: there is little or no documentation about virtual providers and how (or how not) to use them Aug 25 07:40:25 once again, virtual is not special. Aug 25 07:40:31 its just a package Aug 25 07:40:42 ok... Aug 25 07:40:47 that makes sence Aug 25 07:40:57 it was a simple question Aug 25 07:41:02 with a simple answer Aug 25 07:41:19 yes, it was. did someone claim otherwise? Aug 25 07:41:38 no, but you seemed to have a problem with me asking it Aug 25 07:41:43 sorry if it was below you Aug 25 07:41:58 morning Aug 25 07:42:22 it wasnt below me, i simply wanted you to understand that your question was a general one ab out how depends and rdepends work, and was in no way bound to virtual, despite your claims to the contrary Aug 25 07:42:34 now you understand it better, and thats for the good Aug 25 07:42:35 hey hrw, koen Aug 25 07:42:42 hi koen, XorA Aug 25 07:42:44 morning Aug 25 07:42:50 koen: f-b updated again Aug 25 07:42:55 kergoth: not quite... I have been using oe for a couple of months now Aug 25 07:43:03 I understand about RDEPENDS and DEPENDS Aug 25 07:43:18 the question was about virtual packages Aug 25 07:43:24 anyway... I'm clearer on the subject now Aug 25 07:43:24 thanks Aug 25 07:43:42 tkp: virtual/ namespace has sense only in DEPENDS part Aug 25 07:43:56 hrw|work: yes, I understand that now! Aug 25 07:44:50 tkp: no, apparently you do not. you were under the misimpression that RDEPENDS contains bitbake providers. it does not. it contains binary package names from the PACKAGES namespace. Aug 25 07:45:36 but regardless, believe what you want Aug 25 07:46:00 ok, I understand where your comming from Aug 25 07:46:15 * tkp accepts kergoths point Aug 25 07:46:21 god i hate toolchain build failures Aug 25 07:46:27 hate them with a passion Aug 25 07:46:50 morning all Aug 25 07:47:17 morning all Aug 25 07:47:21 hi dirk, rp Aug 25 07:47:35 hey Richard, Marcin Aug 25 07:47:45 morning Aug 25 07:48:18 RP: that "static" fix for poodle, is that also applied to corgi code? Aug 25 07:49:01 so... in one of my bb files, I have RDEPENDS="virtual/libx11", and this works Aug 25 07:49:03 how can that be? Aug 25 07:49:23 XorA: Yes, it should be. I've not tested it though... Aug 25 07:49:36 RP: just about to do the build, Ill test it by lunch Aug 25 07:50:00 tkp: Those packages have an RPROVIDES Aug 25 07:50:20 ahh.. ok Aug 25 07:50:25 tkp: Please don't add virtual RPROVIDES as they're nearly always wrong Aug 25 07:50:41 ok... thats kind of nice tho Aug 25 07:54:53 tkp: hm, that didn't seem to do anything useful Aug 25 07:55:25 this is really weird, task-bootstrap has a number of RDEPENDS which just don't get build when i do a bitbake task-bootstrap Aug 25 08:04:57 http://spyderous.livejournal.com/80869.html Aug 25 08:09:12 good morning all Aug 25 08:09:37 mmp test Aug 25 08:10:02 sorry:) Aug 25 08:10:55 hey koen Aug 25 08:10:57 hey mk Aug 25 08:11:00 hey mikearthur Aug 25 08:11:19 XorA: hey Graeme Aug 25 08:11:20 hey koen, mikearthur Aug 25 08:11:25 hey do13_ Aug 25 08:11:55 do13_: had to patch some of the touch panel stuff so I could get the hsync back easily Aug 25 08:13:30 ok. which files? Aug 25 08:21:44 8664 hrw 0 0 259m 134m 18m S 2.3 17.7 39:17.06 firefox-bin Aug 25 08:21:44 tmiofb.{c,h Aug 25 08:21:47 woops Aug 25 08:21:51 Mem: 775752k total, 763992k used, 11760k free, 81228k buffers Aug 25 08:21:52 Swap: 522072k total, 498076k used, 23996k free, 241204k cached Aug 25 08:21:54 argh Aug 25 08:22:05 tmiofb.{c,h}, tosa.c, tosa_lcd.c Aug 25 08:22:55 hrw|work: that firefox's fault? Aug 25 08:24:14 mikearthur: it takes biggest amount Aug 25 08:24:47 firefox always has been and always will be a memory hog :) Aug 25 08:25:22 mikearthur: and is one of most important tools in my work. Aug 25 08:25:33 mikearthur: why tmiofb? isn't tosa_lcd.c enough? Aug 25 08:25:58 from what I can tell, sadly not, but I'll double check Aug 25 08:26:48 do you have a first version of your patch? Aug 25 08:26:58 doesn't work yet :D Aug 25 08:32:03 hrw|work: need to do some RAM compression :-) Aug 25 08:32:05 oh yeh, do13_, I just checked Aug 25 08:32:43 XorA: probably ;) Aug 25 08:32:59 XorA: in this company I would rather do not get more mem Aug 25 08:33:19 * XorA remembers the good old days of qemm and ram doubler Aug 25 08:33:29 do13_: basically, all I did to tmiofb.c was move some stuff to tmiofb.h, so I could access it externally, and add a function to get_hsynclen Aug 25 08:34:07 * mikearthur remembers the good 'ol days of trying to get DOS to use less than 50k of memory so you could run some game that needed 580k Aug 25 08:34:48 mikearthur: move doom.exe command.com :-) Aug 25 08:34:56 heh Aug 25 08:35:01 never tried that one :) Aug 25 08:35:02 that worked really well Aug 25 08:35:12 then boot from a floppy to fix? Aug 25 08:36:20 nah, what else do you need :-) Aug 25 08:36:34 ~lart git for being so slow Aug 25 08:36:34 * ibot decapitates git conan the destroyer style for being so slow Aug 25 08:38:23 ~lart ibot to see if he has a "anti self destruction" feature Aug 25 08:38:23 * ibot puts ibot to see if he has a "anti self destruction" feature through a wood chipper Aug 25 08:38:41 ok he hasn't one sorry ibot Aug 25 08:41:38 someone know smallest alsa capable mp3/ogg stream player? Aug 25 08:42:19 hrw|work: ogg123 should be capable of doing that... Aug 25 08:42:33 03koen 07org.oe.dev * r68d89a8d... 10/ (1 contrib/feed-browser/update.php): feed-browser: also list number of packages per feed after updating Aug 25 08:43:34 anyone an idea why gcc doesn't like this statement: while(w--) *((unsigned short*)fb)++ = *((unsigned short*)buf)--; ? Aug 25 08:45:01 hrw|work: http://rafb.net/paste/results/3ZG4zg22.html Aug 25 08:45:12 that shows some interesting differences Aug 25 08:47:29 ~curse git fetcher Aug 25 08:47:42 Deltifying 298894 objects. Aug 25 08:47:43 May the fleas of a thousand camels infest your most sensitive regions, git fetcher ! Aug 25 08:48:07 bitbake linux-openzaurus git version = eons in do_fetch, moments in rest Aug 25 08:48:38 koen: build problems shows Aug 25 08:49:34 pgfeller: because its llegal Aug 25 08:49:45 pgfeller: let me guess, asks for modifiable lvalue Aug 25 08:50:16 hrw|work: feed-browser is broken :( Aug 25 08:50:27 koen: which way? Aug 25 08:50:40 hrw|work: http://ewi546.ewi.utwente.nl/tmp/fb/?action=section§ion=base%2Fshell and click on bash Aug 25 08:51:38 will look later Aug 25 08:52:08 the sql looks correctg Aug 25 08:53:20 XorA: you use GNOME or KDE? Aug 25 08:53:27 mikearthur: xfce Aug 25 08:53:49 XorA: Aug 25 08:53:49 ISO C++ forbids cast to n Aug 25 08:53:49 on-reference type used as lvalue Aug 25 08:54:20 pgfeller: basically you have to increment before the cast, not increment the casted value Aug 25 08:54:41 http://ewi546.ewi.utwente.nl/tmp/hrw/fb/?action=details&pnm=g77 Aug 25 08:54:45 works Aug 25 08:54:59 XorA: thanks Aug 25 08:55:31 pgfeller: no probs, I found that nasty in libxine ages ago and had to think why it was busted Aug 25 08:57:11 RP: still get static on corgi, maybe poodle and corgi have different sources. Corgi level of static is defineately related to number of pixels written in fb Aug 25 09:00:53 hrw|work: you're not running the latest code in OE Aug 25 09:01:09 koen: I'm updating now Aug 25 09:02:17 koen: have to merge some things Aug 25 09:02:40 - WHERE (packages.p_name='$package' OR packages.p_provides='$package')- AND feeds.f_name = packages.p_feed + WHERE packages.p_name='$package' AND feeds.f_name = packages.p_feed Aug 25 09:02:44 aargh Aug 25 09:02:51 ~lart appleterm linewrap Aug 25 09:02:52 * ibot holds appleterm linewrap to the floor and spanks him with a cat-o-nine-tails Aug 25 09:08:51 koen: http://ewi546.ewi.utwente.nl/tmp/hrw/fb/?action=details&pnm=bash Aug 25 09:09:18 weird Aug 25 09:09:46 koen: OE version Aug 25 09:09:56 morning all Aug 25 09:10:15 ~hail RP for fixing static noise on corgi. Aug 25 09:10:19 * ibot bows down to RP for fixing static noise on corgi. and chants, "I'M NOT WORTHY!!" Aug 25 09:10:21 hey lrg Aug 25 09:13:26 hey Liam Aug 25 09:13:29 ~seen mickeyl Aug 25 09:13:31 mickeyl was last seen on IRC in channel #oe, 5d 20h 43m 28s ago, saying: 'bbl'. Aug 25 09:14:04 hrw|work: have you update the sqlite db in the meantime? Aug 25 09:14:17 koen: structure is not changed Aug 25 09:14:36 but the code to fill it in has Aug 25 09:14:47 that's the only difference I can think of Aug 25 09:14:49 koen: but fill same data Aug 25 09:16:07 hey Liam Aug 25 09:17:31 koen: and even more - you forgot one field before Aug 25 09:18:26 :) Aug 25 09:18:58 hrw@bitbake:/data/tmp/hrw/fb$ php update.php Aug 25 09:18:58 Updating 3.5.4.1-base: 3485 packages Aug 25 09:25:31 I will probably stop doing builds on work machine during worktime. too much apps has memoryleaks Aug 25 09:25:45 and its hard to work with 15M free ;) Aug 25 09:25:57 hi everybody Aug 25 09:26:11 ran into another error while trying to 'bitbake build-omap1' Aug 25 09:26:21 new console error log -> http://www.rafb.net/paste/results/l6dmJ191.html Aug 25 09:26:29 if anyone can help me out, that would be greatly appreciated Aug 25 09:31:57 hrw|work: no-one needs more than 640K Aug 25 09:32:32 haha Aug 25 09:33:20 koen: 3 files changed, 507 insertions(+), 507 deletions(-) Aug 25 09:34:36 ~seen crofton Aug 25 09:34:55 crofton is currently on #casualti (11h 51m 40s) #ol (11h 51m 40s) #oe (11h 51m 40s) #elinux (11h 51m 40s), last said: 'much of this depends on me remembering stuff I have forgotten'. Aug 25 09:36:03 mickey|sun: ping Aug 25 09:37:37 03hrw 07org.oe.dev * r53526775... 10/ (5 files in 3 dirs): feed-browser: moved functions to include/functions.inc, updater check/create database now Aug 25 09:38:54 n/m ... uncommented the 'depends = uboot' line Aug 25 09:39:15 * bigd0g pets his documentation notebook Aug 25 09:40:45 03pH5 07org.oe.dev * rded2a6ac... 10/ (3 files in 3 dirs): udev-097: drop the MAKEDEV link from links.conf, it is created in the init script Aug 25 09:51:10 mickeyl: alive? Aug 25 09:51:14 kind of Aug 25 09:51:17 hi Aug 25 09:51:20 hey mickeyl Aug 25 09:52:00 mickeyl: python-curses need libncurses5 rdep removal Aug 25 09:52:14 good morning Aug 25 09:53:08 hrw|work: this is a shlibs dependency Aug 25 09:53:12 hrw|work: i can't remove it Aug 25 09:53:21 mickeyl: it gets added automatically Aug 25 09:53:50 mickeyl: and its ncurses not libncurses5 Aug 25 09:54:17 mickeyl: Depends: python-core, ncurses (>= 5.4), libgcc1 (>= 3.4.4), libc6 (>= 2.3.5+cvs20050627) Aug 25 09:54:17 since we include some terminfo into the ncurses package Aug 25 09:54:53 hi mickeyl Aug 25 09:54:56 hey RP Aug 25 09:55:51 RP: sorry for not having responded yet, but your feeling of me being pretty busy was absolutely correct Aug 25 09:56:08 * mickeyl brews a coffee Aug 25 09:56:32 * koen hears rumours about bitbake getting into debian Aug 25 09:57:10 bitbake debian-image would be the coolest thing ever Aug 25 09:57:13 * hrw|work feeds f-b with new packages again Aug 25 09:57:32 apt-get install bitbake would kick ass Aug 25 09:57:46 03pH5 07org.oe.dev * r6651fb70... 10/ (4 files in 3 dirs): initscripts: don't run the devices (devfs) script when udev is active, update for new udev versions Aug 25 09:57:50 mickeyl: np, I understand and guessed as much Aug 25 09:58:03 Updating OpenZaurus 3.5.4 base: 3490 packages Aug 25 09:58:04 lrg: XorA is still hearing static on Corgi :-( Aug 25 09:58:32 koen: includes/config.inc will interest you probably Aug 25 09:58:38 * XorA has also killed Angstrom, time to put oz-unstable on the Z Aug 25 09:59:22 hrw|work: I curse buggy metadata Aug 25 09:59:29 hrw|work: http://www.angstrom-distribution.org/repo/repo.php?action=section§ion=console%2Fnetwork Aug 25 09:59:35 * koen has a look at gpe-edit Aug 25 10:00:30 ! Aug 25 10:00:56 koen: 4 files changed, 138 insertions(+), 85 deletions(-) Aug 25 10:03:05 03koen 07org.oe.dev * r1ee32088... 10/ (1 packages/gpe-edit/gpe-edit_0.32.bb): gpe-edit: insert \n to unbreak metadata Aug 25 10:03:43 hrw|work: i don't really understand. _curses.so dynamically links to libncurses.so.5, so we need that Aug 25 10:04:05 doesn't ncurses provide libncurses5 ? if not, we should make it do that rather than trying to change the shlibs dep Aug 25 10:04:41 03pH5 07org.oe.dev * r916afd44... 10/ (1 packages/linux/handhelds-pxa-2.6_cvs.bb): handhelds-pxa-2.6-cvs: bump version to 2.6.17-hh0 Aug 25 10:05:59 mickeyl: python-*-manifest has NOT needed RDEPENDS entry and thats all Aug 25 10:06:09 -RDEPENDS_python-curses="python-core libncurses5" Aug 25 10:06:09 +RDEPENDS_python-curses="python-core" Aug 25 10:06:25 ah. this i did a couple of weeks ago didn't i? Aug 25 10:06:41 * mickeyl checks Aug 25 10:06:47 I did not see it in my copy of .dev/.oz354x Aug 25 10:07:30 pH5: cant we get handhelds-pxa-2.6_2.6.17-hh0.bb instead? Aug 25 10:07:49 hrw|work: you're right. this change has been dropped somehow :/ Aug 25 10:07:55 i'm going to redo it Aug 25 10:08:15 hrw|work: not 'instead', 'as well' Aug 25 10:08:30 pH5: if I will use SRCDATE=20060612 then handhelds-pxa-2.6_cvs.bb will give me wrong PV Aug 25 10:08:45 koen: good point Aug 25 10:08:46 hrw|work: that's your problem Aug 25 10:09:06 OE metadata is valid for now Aug 25 10:09:16 (albeit annoying) Aug 25 10:09:37 hrw|work: we should rather make a handhelds-pxa-2.6_2.6.16-hh5 Aug 25 10:09:46 hrw|work: actually it will die during do_configure, since it detects a wrong version Aug 25 10:10:27 2.6.17-hh0 won't even compile for hx4700 without the patches I sent to kernel-discuss Aug 25 10:10:44 koen: pull and look at f-b Aug 25 10:11:18 koen: now I see problem which you have Aug 25 10:12:27 koen: did you manage to try NM on your hx4700? Aug 25 10:12:38 florian_kc: not yet Aug 25 10:13:43 koen: ok, you need to start the NetworkManager binary manually at the moment before starting the applet. Aug 25 10:14:09 florian_kc: gpe will/do support wpa and profiles? Aug 25 10:15:44 koen: I know where problem is Aug 25 10:15:45 hrw|work: should both be possiible with it... there are some rough edges at the moment but it starts to become useful. Aug 25 10:17:22 hrw|work: but the quality of the wireless drives makes me headaches Aug 25 10:19:08 florian_kc: does NetworkManager support wlan-ng driver? Aug 25 10:20:58 koen: new version of f-b NEED recreating of database Aug 25 10:21:16 doing that right now :) Aug 25 10:21:30 koen: I did not pushed new one yet Aug 25 10:22:16 do13_: interesting question... if not it might be easy to add there. i guess easier then in gpe-conf Aug 25 10:23:00 public opinion poll: what do you all recommend for an embedded windowing sysem? Aug 25 10:23:36 (i.e. http://linuxdevices.com/articles/AT9202043619.html) Aug 25 10:24:16 directfb Aug 25 10:25:01 gah Aug 25 10:25:08 X11/Kdrive Aug 25 10:25:31 i'm looking for ease of development. both of those qualify? Aug 25 10:26:19 (note: this development is happening on an omap5912osk) Aug 25 10:26:56 bigd0g: how much of memory and flash do you have? Aug 25 10:27:54 32/16 iirc Aug 25 10:28:35 florian_kc: the osk is listed as having 32 ddr sdram / 32 flash Aug 25 10:28:46 florian_kc: that's MBytes Aug 25 10:29:16 bigd0g: what screen size and what kind of application do you expect? Aug 25 10:30:04 mickeyl: screen will be ~3 inches diameter and it's for an embedded personal health monitor Aug 25 10:30:23 mickeyl: (i.e. big text, simple menus, color unnecessary) Aug 25 10:30:24 i see. so just one application at a time, no windows necessary? Aug 25 10:30:32 bigd0g: the advantage of using X is that you have the same API on your PC and on your device. So you can develop on an arbitary PC easily. Aug 25 10:30:49 bigd0g: sounds like a proper task for EFL or Qt/Embedded. Aug 25 10:30:49 mickeyl: right, the application i right will be the only thing the device runs. preferably automatically from boot. Aug 25 10:30:54 hehe Aug 25 10:31:01 s/right/write Aug 25 10:31:49 ok, for a single application and 32mb X should be safe too Aug 25 10:32:08 ok, thanks to: koen, florian_kc, mickeyl -- i've added those to my preferred list. Aug 25 10:32:13 :-) Aug 25 10:32:15 i'll research 'em all and choose accordingly Aug 25 10:32:19 :) Aug 25 10:32:37 bigd0g: your screen is circular? Aug 25 10:33:05 XorA: no, rectangle. 80mm x 61mm Aug 25 10:33:13 XorA: right now, looks like it'll just be b&w Aug 25 10:33:30 XorA: we have to design with diabetic patients in mind, so colors and pretty shapes aren't really an option Aug 25 10:33:46 bigd0g: black/white or greyscale? Aug 25 10:33:58 XorA: you are correct, grayscale Aug 25 10:34:36 ah so worth a windowing system beyond ncurses then Aug 25 10:35:06 well, the device is only meant to run my app. so i'll add ncurses to the list Aug 25 10:36:49 http://ewi546.ewi.utwente.nl/tmp/hrw/fb/?action=details&pnm=hostap-modules is interesting Aug 25 10:37:12 that's neat! Aug 25 10:37:41 mickeyl: org.openembedded.dev/contrib/feed-browser/ Aug 25 10:37:58 hrw|work: couple of those names look familiar ;) Aug 25 10:37:59 cool. is that yours? Aug 25 10:38:13 mickeyl: koen started Aug 25 10:38:18 good work Aug 25 10:38:56 should be usage even in konq/emb 2003 Aug 25 10:39:14 pretty cool, what's that written in? Aug 25 10:39:29 php + sqlite Aug 25 10:40:01 I think it's easy to 'port' it to postgres of *arrgh* mysql Aug 25 10:40:09 heh Aug 25 10:41:09 koen: what for? Aug 25 10:41:35 hrw|work: no idea, but some people insist on using a 'real' db Aug 25 10:41:35 mickeyl: f-b show how fscked sections we have Aug 25 10:43:24 03mickeyl 07org.oe.dev * rc5e8a525... 10/ (1 packages/ipsec-tools/ipsec-tools.inc): ipsec-tools: s/net/network/ Aug 25 10:43:31 03hrw 07org.oe.dev * rdf229e8f... 10/ (5 files in 3 dirs): Aug 25 10:43:31 feed-browser: feeds adding improved, updater more silent Aug 25 10:43:31 - includes/config.inc contain array with all feeds Aug 25 10:43:31 - check_database() adds feeds if they are not present Aug 25 10:43:31 - updater does not show feed URL Aug 25 10:43:37 03mickeyl 07org.oe.dev * r08b0e705... 10/ (1 packages/nfs-utils/nfs-utils_1.0.6.bb): nfs-utils: s/networking/network/ Aug 25 10:43:44 03koen 07org.oe.dev * ra65c738f... 10/ (1 packages/gnome/libgnomecups_0.2.2.bb): libgnomecups: add dep on libgnomeui Aug 25 10:43:50 03hrw 07org.oe.dev * rd9256d2a... 10/ (4 files in 3 dirs): Aug 25 10:43:50 feed-browser: FIX: feed structure was changed - database NEED to be rebuilt Aug 25 10:43:50 - package.p_feed is now int -> feeds.f_id Aug 25 10:45:54 hmm Aug 25 10:46:02 we're still missing about 200 people on the new list Aug 25 10:46:09 do you think we should send a reminder? Aug 25 10:46:16 (whoever this 'we' may be ;) Aug 25 10:47:14 mickeyl: you are royalty, in UK royalty may address themselves as we :-) Aug 25 10:47:33 * mickeyl bows to XorA Aug 25 10:47:48 * XorA suspects that is backwards Aug 25 10:47:57 heh Aug 25 10:48:12 the poor peasants like us normally bow to the royalty :-) Aug 25 10:48:44 heh. sometimes the roalty needs to bow to encourage the lower beings to continue their struggle, right? ;) Aug 25 10:48:49 hm, with my current build mb-applet-menu-launcher dies with a BadAccess X Error (request MIT-SHM, X_ShmAttach), sometimes directly, sometimes after I click the button. This does not happen if I start it as root. Aug 25 10:48:57 Is this a permission error? Aug 25 10:49:00 mickeyl: I thought that was what whips were for :-) Aug 25 10:49:18 XorA: hehe, oh right. _that_'s what they're for :D Aug 25 10:49:33 hrw|work: Warning: sqlite_query(): table feeds has no column named f_id in /var/www/angstrom/repo/includes/functions.inc on line 93 Aug 25 10:49:40 mickeyl: unless Eurotrash is watched Aug 25 10:49:49 *nod* Aug 25 10:50:01 pH5: angstrom seems to have shm issues Aug 25 10:50:25 hrw|work: the zsafe packaging errors on openzaurus ml -- do you look into or should i? Aug 25 10:50:26 hrw|work: after I deleted all dbs Aug 25 10:50:48 koen: I noticed.. It seems to be working fine as root, though. Aug 25 10:51:11 pH5: it's also working fine if you start it the second time.... Aug 25 10:53:46 hrw|work: it seems you forgot to commit the check_database() updates Aug 25 10:54:49 koen: 25 12:59 < CIA-9> feed-browser: FIX: feed structure was changed - database NEED to be rebuilt Aug 25 10:54:58 koen: ops Aug 25 10:55:09 mickeyl: oops... 199, i just subscribed Aug 25 10:55:40 koen: yep Aug 25 10:56:33 mickeyl: you? Aug 25 10:58:49 koen: pushed Aug 25 11:00:04 03hrw 07org.oe.dev * r985cd959... 10/ (1 contrib/feed-browser/includes/functions.inc): feed-browser: forgotten structure change Aug 25 11:00:05 thanks Aug 25 11:00:46 hrw|work: do you know why the pkgconfig in oz354x doesn't install the included pkg.m4? Aug 25 11:01:26 florian_kc: dont know Aug 25 11:02:21 florian_kc: excellent ;) Aug 25 11:02:31 hrw|work: ok. i'll add this and build a tree to find out if it causes trouble Aug 25 11:02:43 ok Aug 25 11:02:49 hrw|work: hmm thinking about it... we don't have zsafe at all. all we have is opie-zsafe and that was your baby wasn't it? Aug 25 11:03:33 mickeyl: we had zsafe and opie-zsafe and then zsafe got removed Aug 25 11:05:16 mickeyl: http://openembedded.bkbits.net:8080/openembedded/cset@1.755 Aug 25 11:05:29 http://openembedded.bkbits.net:8080/openembedded/cset@1.1706.270.1 == removed zsafe because opie-zsafe is being updated regularly. Aug 25 11:07:23 quick question of which i couldn't really find the answer on the wiki, strange enough: where do i donate the $$$? Aug 25 11:08:20 that's a good question Aug 25 11:08:25 * koen looks at mickeyl Aug 25 11:09:07 i'm not sure if i can talk my manager into donating, but i sure will try Aug 25 11:09:56 we don't have a common project account yet, so chose one guy you think is responsible Aug 25 11:10:08 we try to get a common project account done at OEDEM Aug 25 11:10:11 later this year Aug 25 11:10:12 or the hosting provider Aug 25 11:10:42 i may need a bit of money to do flyers and posters for FOSDEM07 Aug 25 11:10:47 giel: did you do some work on the new LPC3xxx chips? Aug 25 11:11:07 * koen gets reminded to print shirts Aug 25 11:11:10 okay, well, i'm still developing anyway, we should have a working demo end of september, if that succeeds and is made possible by OE, I think i can convince people to hand out cash Aug 25 11:11:16 mickeyl: OEDEM wiki Aug 25 11:11:42 koen: LPC3xxx is external name i think, i don't really know the mapping between the numbers used externally and internally Aug 25 11:11:55 koen: i know... /me writes on TODO for weekend Aug 25 11:11:56 giel: the arm926ej-s with VFP Aug 25 11:12:09 i am working with such a thing right now Aug 25 11:12:22 i'm using it, not developing the cpu itself Aug 25 11:12:28 cool Aug 25 11:12:50 for the moment i'm not using the VFP tho, CSL gave me nothing but weird problems Aug 25 11:13:16 * koen says nothing about csl toolchains Aug 25 11:13:57 oh, i'm not saying it's the toolchains per se, it's just that I first want to have working stuff with a normal toolchain Aug 25 11:14:52 e.g., the 3.4.0-vfp csl, i think that's 2004q1, just errors out with internal compiler errors from time to time Aug 25 11:15:31 giel: and can give non-working binaries Aug 25 11:15:56 iirc the maemo people are happy with the 2005q3 toolchain Aug 25 11:16:09 arm926ej-s without vfp Aug 25 11:16:59 anyway, i have a happily working 3.4.4 with glibc 2.3.5+cvs toolchain now Aug 25 11:20:50 03koen 07org.oe.dev * r25188d2b... 10/ (1 packages/xorg-xserver/xserver-kdrive_git.bb): xserver-kdrive git: fix packaging so Security Policy ends up in the rootfs Aug 25 11:23:47 koen: funny, I was just looking into this for X11R7.1 Aug 25 11:23:56 koen: does the git version build Xsdl, too? Aug 25 11:24:09 pH5: if you build sdl before kdrive, yes Aug 25 11:25:44 http://rafb.net/paste/results/4J9aHr37.html Aug 25 11:27:55 koen: own config.inc made? Aug 25 11:28:06 yes Aug 25 11:28:45 koen: we need to add kind of welcome page too Aug 25 11:37:18 so Xsdl on sdl-opie is a possability :-) Aug 25 11:37:33 sdl-qpe I mean Aug 25 11:44:48 XorA: X + qvfb Aug 25 11:48:59 hrw|work: where are the sources for OZ3.5.4.x? Aug 25 11:53:56 hrw|work: if they are on ewi I'd like to use them as a tarball stash :) Aug 25 11:54:25 koen: ewi Aug 25 11:54:32 koen: /3.5.4/sources/ Aug 25 11:54:41 * koen adds Aug 25 11:55:19 koen: opie + x/qt + qvfb + Xsdl Aug 25 12:09:47 koen: http://ewi546.ewi.utwente.nl/tmp/hrw/fb/?action=details&pnm=libsdl-1.2-0 - this show that we need to change details view Aug 25 12:11:30 yes Aug 25 12:11:33 and http://www.angstrom-distribution.org/repo/repo.php?action=details&pnm=kernel-module-aes Aug 25 12:11:37 that shows about the same Aug 25 12:12:14 details per packages and the links below point to the details for the other archs Aug 25 12:13:03 koen: similiar to http://packages.debian.org/monotone Aug 25 12:13:21 yes Aug 25 12:13:35 although I use the ubuntu one because the debian one is down too often Aug 25 12:19:12 XorA: I just got that fontconfig related horribly slow starting of apps, too :-/ Aug 25 12:22:02 koen: http://ewi546.ewi.utwente.nl/tmp/hrw/fb/?action=details&pnm=libsdl-1.2-0 Aug 25 12:22:47 koen: a bit to hardcore Aug 25 12:22:53 hrw|work: I wouldn't make it a bullet list, but divs with alternating background Aug 25 12:24:11 from semantic point of view it is a list Aug 25 12:25:16 it is Aug 25 12:26:45 lack of usability Aug 25 12:35:54 pH5: :-( Aug 25 12:36:31 pH5: I miss the days of 8x8 bitmap fonts that drew fast :-) Aug 25 12:47:33 grr, I am having preferred version issues Aug 25 12:49:34 03mickeyl 07org.oe.dev * r62a2d9a8... 10/ (1 packages/libvncserver/libvncserver_0.8.2.bb): libvncserver: depends on libxext Aug 25 12:49:39 03mickeyl 07org.oe.dev * r9372040a... 10/ (5 files in 3 dirs): contrib/python: add the script that I use to generate the python manifest file Aug 25 12:51:00 yay! more contrib Aug 25 12:51:49 mickeyl must be procrastinating Aug 25 12:51:59 thats a lovely word Aug 25 12:52:15 03mickeyl 07org.oe.dev * re3f12501... 10/ (1 contrib/python/generate-manifest.py): contrib/python/generate-manifest.py: update to latest version Aug 25 12:52:18 what does it mean? Aug 25 12:52:25 03mickeyl 07org.oe.dev * r9c1afd1e... 10/ (1 packages/python/python-2.4.3-manifest.inc): python-manifest: remove explicit rdependency on libncurses5 Aug 25 12:52:32 mickeyl: delaying stuff Aug 25 12:52:44 I have a line PREFERED_PROVIDER_virtual/kernel = "linux-omap1" Aug 25 12:53:04 mickeyl: over here we say 'Study avoiding behaviour' Aug 25 12:53:10 and a PREFERED_VERSION_kernel = "2.6.12-rc2" Aug 25 12:54:17 but it inists on building the "2.6.17-omap1" version Aug 25 12:54:56 PREFERED_VERSION_linux-omap1 = "2.6.12-rc2" Aug 25 12:55:04 ok Aug 25 12:55:06 thanks Aug 25 12:55:21 so many details, so little time Aug 25 13:24:07 morning Aug 25 13:26:16 ~lart SoC page Aug 25 13:26:16 * ibot slams SoC page against a large cement Tux Aug 25 13:27:31 I hate pages where I have to look into source to check where are links... Aug 25 13:32:35 could I make a donation to oe? Aug 25 13:32:45 that would make my life easier :) Aug 25 13:32:50 donate a flipping serial cable for this tosa Aug 25 13:33:15 mikearthur: you do not have serial cable? Aug 25 13:33:22 nope, none came with it :s Aug 25 13:33:34 its needs a funny "I/O port" connector Aug 25 13:33:57 I know Aug 25 13:33:59 I have one Aug 25 13:34:33 trisoft want 65 EUR for one. I do not know will they want to donate one... its not 10 eur hostcable. Aug 25 13:35:03 in fact, wolfson will get one Aug 25 13:35:18 mikearthur: you work in wolfson too? Aug 25 13:35:26 mikearthur: get two :) Aug 25 13:35:26 http://www.yoss.de/cgi-bin/yosst.htm?shop=10&direkt=ja&language=2&jump=CE-170TS Aug 25 13:35:28 technically ;) Aug 25 13:35:37 I'm lrg's student slave Aug 25 13:35:40 ah Aug 25 13:35:58 mikearthur: work in chains and have to ask if want to leave workplace? Aug 25 13:36:15 and loose at pool? Aug 25 13:37:10 hrw|work: leaving the workplace? isn't that what happens when I die? Aug 25 13:37:12 and he's can't even make a decent coffee Aug 25 13:37:14 koen: beat him today Aug 25 13:37:41 lrg: badly choosed then.. Aug 25 13:38:08 lrg: he need the right training Aug 25 13:38:42 mikearthur: only if you will be lucky Aug 25 13:40:24 hrw|work, do13: I might let him come to FOSDEM next year. Aug 25 13:41:06 fosdem07.. I have to be there Aug 25 13:43:30 hrw|work: yes! Aug 25 13:44:16 mikearthur: make sure that gmp suit is clean when you leave :-) Aug 25 13:44:50 koen: will do what can to be there Aug 25 13:44:57 XorA: you can clean it when you need to wear it Aug 25 13:45:14 (pwned) Aug 25 13:46:00 hrw|work: I think for £40 I'll just live with it atm :s Aug 25 13:47:00 mikearthur, lrg, XorA: none of you has serial cable for Z? strange Aug 25 13:47:34 hrw|work: I have the bits to make one Aug 25 13:47:41 hrw|work: one day I will have the time Aug 25 13:47:58 hrw|work: then I shall buy a computer with a serial port :-) Aug 25 13:48:11 XorA: I have bits for over two years.. Aug 25 13:48:21 XorA: you will have serial port soon. Aug 25 13:48:21 XorA: pl2303 usbserial are cheap Aug 25 13:49:54 hrw|work: RP kernel is too damn good to need serial cable :-) Aug 25 13:50:10 hi all; does anyone know where "gnu-configize" is documented (or implemented)? Aug 25 13:52:06 pgfeller: its packages/gnu-config Aug 25 13:53:45 hrw: thanks - found it (now I'll have to find out what it does :-) ... Aug 25 13:53:59 koen: http://www.openembedded.org/repo/org.openembedded.dev/contrib/feed-browser/index.php ;( Aug 25 13:55:06 heh Aug 25 13:55:22 I suspect someone needs to add a .htaccess Aug 25 13:57:33 then do it Aug 25 13:59:20 or someone else Aug 25 14:03:32 * RP has a serial lead funnily enough. Shame it only works properly with the c7x0 though :-( Aug 25 14:03:45 (its a serialIO one) Aug 25 14:04:17 serialIO works on the tosa too Aug 25 14:04:39 sharp with collie poodle tosa c7x0 Aug 25 14:05:20 It doesn't work with cxx00 (or poodle but that could be a setup issue in locomo) Aug 25 14:10:48 pgfeller: btw, the -x11 and -qpe versions of libsdl must match, so you can't add 1.2.11-qpe without adding 1.2.11-x11 Aug 25 14:11:42 koen: really? iirc you added newer -x11 one... Aug 25 14:11:53 hrw|work: after talking to mickeyl Aug 25 14:12:28 ah Aug 25 14:12:45 koen: oh ... well - initialiy I just wanted to build libvncserver against libsdl-qpe :-/ Aug 25 14:14:28 I have a libsdl-x11-1.2.11 here, but it doesn't compile yet: Aug 25 14:14:28 "libtoolize: invalid AC_CONFIG_AUX_DIR syntax: AC_CONFIG_AUX_DIRS($srcdir/build-scripts) Aug 25 14:15:33 ph5: had the same problem with qpe version ; if I override do_configure () with do_configure() { gnu-configize oe_runconf( Aug 25 14:15:54 ph5: but I'm still in the process of learning what this actually did ;) Aug 25 14:17:15 maybe it need autoconf 2.60? Aug 25 14:18:29 ph5: missed to say that it worked with the override & not without Aug 25 14:21:10 pgfeller: what override? Aug 25 14:22:02 Hi! Aug 25 14:22:04 ph5: if I define do_configure () { gnu-configize oe_runconf} in the .bb it builds Aug 25 14:22:26 I wonder, is it possible to override IMAGE_FSTYPES and IMAGE_ROOTFS_SIZE on per-image basis? Aug 25 14:23:50 psokolovsky: yes Aug 25 14:24:10 hrw|work: any pointer to example? Aug 25 14:24:12 psokolovsky: look at packages/images/pivot*image.bb Aug 25 14:24:21 hrw|work: Thanks! Aug 25 14:24:45 psokolovsky: in .oz354x branch Aug 25 14:24:51 ok Aug 25 14:25:21 http://www.openembedded.org/repo/org.openembedded.oz354x/packages/meta/pivotboot-image.bb Aug 25 14:27:53 hrw|work: I see it's overriden based on machine name. What would be syntax based on image name? Or just set "IMAGE_FSTYPES" in .bb and that will override local.conf setting? Aug 25 14:28:37 psokolovsky: pivotboot-image.bb is image Aug 25 14:28:55 OE does not have PN in overrides Aug 25 14:29:26 so image need atleast some code for checking do you want something extra Aug 25 14:30:09 hrw|work: Aha. I'm just concerned that local.conf defines IMAGE_FSTYPES with "=", not "?=". Aug 25 14:31:34 psokolovsky: local.conf is read before machine/distro so it usually does not have other conf before itself Aug 25 14:31:43 pgfeller: with gnu-configize it works here, too. I wonder if the syntax of AC_CONFIG_AUX_DIRS changed? Aug 25 14:33:08 erk, oz-unstable image is now bigger than 64M Aug 25 14:33:50 ph5: no idea ... took me hours to understand just the concept of the gnu build mechanics ... Aug 25 14:34:27 hrw|work: Ok, I see that 'IMAGE_FSTYPES = "ext2.gz" Aug 25 14:34:27 IMAGE_ROOTFS_SIZE_ext2.gz = "12288" Aug 25 14:34:27 ' in .bb file does exactly what we need, overriding values in local.conf/bitbake.conf. Thanks for help! Aug 25 14:34:45 np psokolovsky Aug 25 14:34:48 XorA: o.. Aug 25 14:36:36 ph5: I had a look into the autotools class - but the magic performed there was to much for me to understand. Do you know what gnu-configize is for (when should it be employed)? Aug 25 14:37:13 gnu-configize copies the default config.sub and config.guess into the srcdir Aug 25 14:37:36 the configure.in in sdl wants to tell autoconf to get those files from build-scripts/ instead, but that somehow doesn't work out Aug 25 14:38:23 pH5: do you know what's the status on the acx drivers? Aug 25 14:40:34 koen: not sure. I can load the modules, connect to my ap, surf the web and unload the modules again. Aug 25 14:41:01 same here Aug 25 14:43:45 Those Xshm issues worry me far more currently. shm_attach in kdrive somehow decides that I'm not allowed to use shm, but I don't understand why. Aug 25 14:45:15 pH5: XorA had some more shm issues iirc Aug 25 14:46:03 playing with xauth a little bit helped somewhat but not quite. Aug 25 14:52:16 03florian 07org.oe.dev * rcedc9263... 10/ (1 packages/networkmanager/networkmanager_0.6.4.bb): networkmanager: depend on libglade Aug 25 14:55:46 pgfeller: the solution is simple: it's a bug in the libtoolize script that doesn't know about AC_CONFIG_AUX_DIRS. by overwriting do_configure you just make sure that libtoolize is not run. Aug 25 14:56:27 03florian 07org.oe.oz354x * r26de4ae1... 10/ (1 packages/pkgconfig/pkgconfig_0.15.0.bb): pkgconfig: Install updated pkg.m4. Aug 25 14:57:06 ph5: is this bug in the autotools.class of oe, or in libsdl? Aug 25 15:03:50 XorA: what make is your PDA Aug 25 15:03:55 I meant model Aug 25 15:03:55 I don't know. Could also be libtool. Aug 25 15:03:57 mikearthur: Zaurus Aug 25 15:04:03 mikearthur: C860 Aug 25 15:04:17 mikearthur: or HTC Wizard Aug 25 15:10:28 XorA: are there any Sharp clamshells with inbuilt wifi? Aug 25 15:10:35 mikearthur: no Aug 25 15:10:49 damn Sharp, damn them to hell Aug 25 15:10:49 mikearthur: only the tosa has bt/wifi Aug 25 15:11:28 but no real keyboard :s Aug 25 15:11:43 get a htc universal :) Aug 25 15:12:00 koen: only the tosa *can have* bt/wifi builin Aug 25 15:12:17 some tosa models lack bt, some lack wifi, some lack both Aug 25 15:12:44 people should have gotten suspicious when sharp didn't use their own arm chips Aug 25 15:12:45 koen: better sectionlist pushing.. Aug 25 15:13:44 * lrg would like a pda with the mx31 - ARMv6, with VFP and hardware MPEG, 3D Aug 25 15:13:56 lrg: yes! Aug 25 15:14:08 lrg: no more alignment errors and softfloat! Aug 25 15:14:35 koen: whats the htc universal? Aug 25 15:14:38 koen: so far it is 2 * faster than pxa27x Aug 25 15:14:58 mikearthur: the clamshell phone thingy Aug 25 15:15:39 koen: my dad has one, doesn't run linux though, eh? Aug 25 15:15:45 mikearthur: can run Aug 25 15:15:50 mikearthur: conf/machine/htcuniversal.conf Aug 25 15:15:55 mikearthur: but then it is only pda.. not phone Aug 25 15:16:33 hrw|work: would it be eventually possible for it to be oth? Aug 25 15:16:47 maybe one day... Aug 25 15:17:01 hrw|work: If I'm good... :D Aug 25 15:18:49 03hrw 07org.oe.dev * r28a2710b... 10/ (1 contrib/feed-browser/includes/functions.inc): feed-browser: much better sectionlist (rewrite to be recursive anyone?) Aug 25 15:21:43 koen: feeds need flag 'upgrades = 0/1' so we will know do we have to read them or not on update Aug 25 15:21:58 that's be nice Aug 25 15:22:14 I'm feeding my f-b now Aug 25 15:22:25 section = strtolower(section) Aug 25 15:22:57 opie-sh depend on opie-sh Aug 25 15:23:13 thejapa: hi and bye - I have to go Aug 25 15:23:28 hrw, k Aug 25 15:23:29 hurry Trolltech up with the greenphone and ROAD with their phone Aug 25 15:23:35 thejapa: will look at your source during next week Aug 25 15:23:35 and give OE free samples :-) Aug 25 15:23:46 hrw, no hurry anymore :) Aug 25 15:24:20 cu guys - be back on Monday - tomorrow wedding. Ania's brother is leaving world of free people... Aug 25 15:24:49 cu Marcin Aug 25 15:26:59 http://www.angstrom-distribution.org/repo/repo.php?action=details&pnm=angstrom-base-wifi Aug 25 15:27:09 the new section stuff looks pretty neat Aug 25 15:27:28 koen: Can you remind me which toolchains work with arm mmx please? Aug 25 15:28:06 gcc4.1.1/glibc2.4 EABI Aug 25 15:28:16 RP: DISTRO-angstrom-2006.9 ;) Aug 25 15:28:26 I'm using iwmmxt since gcc3.3 ;) Aug 25 15:28:31 koen: The csl ones are broken in that regard? Aug 25 15:28:54 csl toolchains are broken Aug 25 15:29:11 koen: true :-/ Aug 25 15:29:29 koen: I presume hardfloat gcc 3.4.4 with mmx also doesn't work for some reason? Aug 25 15:29:39 it doesn't Aug 25 15:29:46 intel put iwmmx on CP0 iirc Aug 25 15:30:01 which overlaps with FPA instructions Aug 25 15:30:37 koen: right. madness :-/ Aug 25 15:31:21 Hello. Aug 25 15:31:43 hey sirfred Aug 25 15:32:05 RP: Hey, I've found an interesting thing on the IMAGEON, I don't know if you were aware of it. Aug 25 15:32:23 hi sirfred Aug 25 15:32:52 RP: Register fields that seemed duplicated, like, for example DP_GUI_MASTER_CNTL.rop3 and DP_MIX.rop3, are internally the same value. Aug 25 15:33:34 sirfred: Interesting. You discovered that by poking bits? Aug 25 15:33:39 RP: If you change one of them, the other one gets changed. Aug 25 15:33:52 RP: Yes, I just changed the first, and the latter gets changed . Aug 25 15:33:54 Thats interesting to know... Aug 25 15:34:14 RP: Just a way to optimize the number of register writes. Aug 25 15:34:26 RP: But it allows me a lot to understand how the thing works. Aug 25 15:34:37 s/allows/helps Aug 25 15:34:49 sirfred: Every little piece of knowledge helps :) Aug 25 15:35:10 RP: Well, it's just that I've just discovered it, and wanted to tell it to someone. I know you're too busy. Aug 25 15:36:18 sirfred: I know what its like. It is really good when you find something like that :) Aug 25 15:36:36 RP: I've worked a little in the wiki, slightly modifying a script Mardy gaves me to transform the header into wiki format tables. This is how it is going: Aug 25 15:36:37 http://libw100.sourceforge.net/index.php/Data_Path#DP_GUI_MASTER_CNTL Aug 25 15:37:55 sirfred: with the debug bit will it also support 565 destinations? Aug 25 15:38:16 What, exactly? Aug 25 15:38:30 Alpha blending? Yes. Aug 25 15:38:36 sirfred: Just looking at Aug 25 15:38:36 gmc_dst_datatype Aug 25 15:38:49 RP: Yes. Aug 25 15:39:09 RP: That needs to be documented, but a lot of registers are still missing. Aug 25 15:39:15 RP: It is a slow work. Aug 25 15:39:52 RP: The bad thing is that I didn't found a way to support YUV surfaces with this raster operations. Aug 25 15:40:33 sirfred: Slow but appears steady now :) Aug 25 15:40:47 sirfred: Its a shame YUV can't work :-/ Aug 25 15:41:01 RP: Who knows, perhaps another debug bits... :) Aug 25 15:41:13 sirfred: you never know :) Aug 25 15:41:34 Too much debug registers. I expect most of their bits to do nothing. :) Aug 25 15:42:59 sirfred: Quite likely. I wish we could get docs on them! Aug 25 15:45:32 Have to leave now. bbl. Aug 25 15:58:15 03pfalcon 07org.oe.dev * r8f215c16... 10/ (1 conf/machine/h4000.conf): h4000: Use modutils 2.6. Aug 25 16:01:56 hmm sould I have Aug 25 16:02:15 DEPENDS = "u-boot" Aug 25 16:02:20 or "uboot" Aug 25 16:07:07 a+ Aug 25 16:08:21 03yuecelm 07org.oe.dev * r5d6540ad... 10/ (1 packages/initscripts/initscripts-1.0/slugos/devices.patch): initscripts-1.0/slugos/devices.patch: adapt patch to the current devices file Aug 25 16:21:15 Crofton: how's the u(-)boot recipe called? Aug 25 16:21:33 not quite sure what you mean Aug 25 16:21:42 there is a directory packages/uboot Aug 25 16:21:57 that contains uboot*bb and u-boot*bb files Aug 25 16:22:04 great..... Aug 25 16:22:14 thus my question :) Aug 25 16:22:27 I guess you'd want u-boot Aug 25 16:22:33 since you added that :) Aug 25 16:22:37 Yeah :) Aug 25 16:22:50 but I patterned it after the existing one I used Aug 25 16:23:19 only the cvs and pxa255 ones are uboot Aug 25 16:24:03 I need to push some changes to an omap kernel Aug 25 16:24:09 so it actually works Aug 25 16:24:23 and need the u-boot dependency in there Aug 25 16:28:32 03yuecelm 07org.oe.dev * r5cca4bb7... 10/ (3 files in 2 dirs): Aug 25 16:28:32 avetanabt: add version 20060413 Aug 25 16:28:32 - cvs version: updated path to CVS repository, DEFAULT_PREFERENCE = "-1" Aug 25 16:29:49 hmmm, libungif fails with "CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include". is there a fix available? Aug 25 16:30:02 (i'm not up-to-date, so sorry if there is something in .dev) Aug 25 16:31:35 tmbinc: patch it to stop doing -I/usr/include Aug 25 16:31:54 x29a: ping Aug 25 16:32:06 koen, is it ok if I use DEPENDS="u-boot" ? Aug 25 16:32:10 tmbinc: this was a quick hack to gcc Aug 25 16:32:12 Crofton: sure Aug 25 16:32:16 ok Aug 25 16:32:24 tmbinc: if we see /usr/include /sw/include /usr/local/include it will abort Aug 25 16:32:40 will commit some stuff after lunch ... Aug 25 16:33:27 zecke: yeah i've seen this. So, there is no fix for libungif yet, right? Aug 25 16:33:38 then i'll try to fix it Aug 25 16:35:18 zecke: I'll start preparing .dev for packaged staging after next week Aug 25 16:54:10 mr_nice_slacker1: pong Aug 25 16:54:30 mr_nice_slacker1: im sorry, i just flashed hh.org bldr but weve got company, so im kinda busy with my brothers Aug 25 16:54:34 ill talk to you later Aug 25 17:09:29 koen: hmm, I'm doing the survey now Aug 25 17:09:38 koen: I wonder if you completed the project Aug 25 17:09:46 koen: or if it wasn't realizable Aug 25 17:14:49 Can somebody quickly explain to me the meaning of the colours and stuff on http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=OpenEmbeddedBuilds ? Just being curious. Aug 25 17:15:36 Laibsch: Red == Failed build Aug 25 17:15:39 Laibsch: Green success Aug 25 17:15:45 Laibsch: yellow == work in progress Aug 25 17:17:44 03yuecelm 07org.oe.dev * r641b5fa9... 10/ (1 packages/subversion/subversion_1.3.2.bb): Aug 25 17:17:44 subversion: add (disabled) version 1.3.2 Aug 25 17:17:44 - DEFAULT_PREFERENCE = "-1" due to errors on stage phase Aug 25 17:17:44 - bug report follows soon Aug 25 17:17:57 zecke: Thanks. And what about the black stuff up on the top which seems to be an attribute of the build machine? Aug 25 17:18:31 What about grey? Unknown outcome? Aug 25 17:21:15 Laibsch: black probably means needs attention Aug 25 17:21:31 Laibsch: gray means, build has not finished but a new one started Aug 25 17:21:47 OK, I see. Thanks. Aug 25 17:21:56 Compiling away. Aug 25 17:23:38 zecke: a few weeks ago I'd have said it was realizable Aug 25 17:23:49 zecke: after thinking some more it's a lot harder Aug 25 17:28:22 How does bootstrap image knowhow big to make the jffs image? Aug 25 17:30:57 Crofton: it doesn't Aug 25 17:31:28 Crofton: it will happily create a gb sized jffs2 if you include a lot Aug 25 17:31:47 how do I make it the "right" size? Aug 25 17:31:59 -p ? Aug 25 17:32:39 I feel like I am missing something here Aug 25 17:32:53 I have a board, with a "partition" of a certain size Aug 25 17:33:04 I want to make a jffs file system to go in it Aug 25 17:33:19 * Crofton my not know what he is doing .... Aug 25 17:33:21 most of the times you don't have to pad it Aug 25 17:33:39 works by magic? Aug 25 17:33:47 you can happily flash a smaller jffs2 and it will automagically use the free space Aug 25 17:33:53 ok Aug 25 17:34:00 but trouble if you go over? Aug 25 17:34:08 yes Aug 25 17:34:11 ok Aug 25 17:34:23 that is enough for me to move forward Aug 25 17:39:36 can bitbake generate an ext3 image? Aug 25 17:40:42 tkp: pass -j as an extra param for ext2? Aug 25 17:41:02 hi does anyone want to help port (J2ME)CLDC + MIDP2.0 to zaurus ? Aug 25 17:41:03 koen: really? cool... I'll give it a go Aug 25 17:47:51 koen: you know where the actual bbimage command is defined? Aug 25 17:48:12 classes/ ? Aug 25 17:48:44 hmm... thats what I thought Aug 25 17:48:51 dont see it there tho :( Aug 25 17:49:40 bitbake/bin/bbimage Aug 25 17:51:57 even better - oe-stv/org.openembedded.dev/conf/bitbake.conf Aug 25 17:53:03 03crofton 07org.oe.dev * rd062b97f... 10/ (1 conf/machine/omap5912osk.conf): Aug 25 17:53:03 omap5912conf : Fix preferred version variables Aug 25 17:53:03 * use linux-omap1_2.6.12-rc2 for now Aug 25 17:53:03 * prepare to use linux-omap1_2.6.17-omap1 when it works Aug 25 17:53:09 03crofton 07org.oe.dev * r6e527644... 10/ (1 conf/distro/openomap.conf): distro/openomap.conf : Update toolchain versions Aug 25 17:53:14 03crofton 07org.oe.dev * r910e407e... 10/ (1 packages/linux/linux-omap1_2.6.12-rc2.bb): Aug 25 17:53:14 linux-omap1_2.6.12-rc2.bb : Fixes so it builds reliably Aug 25 17:53:14 * add dependency on u-boot Aug 25 17:53:14 * Fix do_deploy to massage kernel for use on omap5912osk board Aug 25 17:53:19 EXTRA_IMAGECMD :) Aug 25 18:01:41 03koen 07org.oe.dev * rd6f3bb86... 10/ (1 classes/base.bbclass classes/package.bbclass): classes: run do_package before do_stage so we can populate staging with package if we want Aug 25 18:19:55 Is there a way to make do_unpack extract only one directory from a tarball, instead of the whole thing? Aug 25 18:23:19 mattcamp: Why would you want to do that? Aug 25 18:23:28 Just ignore the stuff you do not need. Aug 25 18:28:32 hi Aug 25 19:26:56 wow screaming to "the doors" makes fun Aug 25 19:27:23 * chouimat is listening to "The Legacy" by Iron Maiden on A Matter of Life and Death [amaroK] Aug 25 19:27:34 hello all Aug 25 19:27:59 chouimat: scream to "Been Down So Long", and makes one happy Aug 25 19:28:05 it makes one happy Aug 25 19:28:48 okey to use 0.29 monotone for OE? Aug 25 19:28:58 zecke: hehe ;) Aug 25 19:29:19 Kristoffer: didn't know that one was released Aug 25 19:29:23 Kristoffer: I think it should be fine Aug 25 19:29:33 zecke, oki thx Aug 25 19:29:49 zecke: I'm happy today ... new maiden album and my laptop is back from the dead ... Aug 25 19:31:33 chouimat: take care mate Aug 25 19:33:58 zecke: I will ... and I think that I will leave now for the local pub ... that way I will be able to get half priced appertizers :) Aug 25 19:35:28 Kristoffer: http://lists.gnu.org/archive/html/monotone-devel/2006-08/msg00373.html Aug 25 19:35:59 doh Aug 25 19:36:01 zecke, thx man Aug 25 19:37:51 Kristoffer: try the static binary first Aug 25 19:38:04 njs: ever considered using dietlibc and uclibc and linking that statically? Aug 25 19:39:37 zecke, will do Aug 25 19:39:49 I will build a mtn 0.29 as well (OSX/i386) Aug 25 19:45:42 Kristoffer: it was a wrong compilation, so mtn 0.29 should be fine Aug 25 19:45:51 aha oki Aug 25 19:46:23 Btw, just for interested. I use gentoo and there 0.19 is the stable one. Any logical reason for this? Looks like monotone has come a long way since 0.19 Aug 25 19:47:23 Kristoffer: old cruft is the only reason I can think of Aug 25 19:51:29 zecke: 0.29 works for me Aug 25 19:53:28 it's running on monotone.oe.org for a while now iirc Aug 25 19:54:13 anyway ... see you later I'm heading for the pub ... Aug 25 19:57:25 Kristoffer: because no-one has removed the ~x86...for some reason. If you keyword it it will build a much newer version (0.25 or 26 I think) Aug 25 19:57:45 Kristoffer: I created an overlay and just copied the newest file for 0.29 and it builds and works fine Aug 25 21:50:22 nite Aug 25 22:36:40 anyone messed w/ TT's new phones? Aug 25 23:25:40 hm, what's wrong when "mount -a" mounts already mounted points again? ("/var" in this case) Aug 25 23:30:43 it shouldnt be able to without -o remount Aug 25 23:31:21 probably fails silently Aug 25 23:34:09 no it doesn't fail, it mounts it over the old mountpoint Aug 25 23:34:21 because it's tmpfs Aug 25 23:39:21 i'll have to try that Aug 25 23:42:04 busybox 1.01 here Aug 26 00:16:28 Are OpenSlug and OpenZaurus endian-compatible? I'd like to keep just a single package tree for both if possible. Aug 26 00:26:21 03tmbinc 07org.oe.dreambox * rcee4d59c... 10/ (1 packages/libungif/libungif_4.1.3.bb): libungif: compile without X11 support to fix bad includes Aug 26 00:26:27 03tmbinc 07org.oe.dreambox * r54819faa... 10/ (1 packages/tuxbox/tuxbox-libs.bb): tuxbox-libs: add proper PACKAGES_DYNAMIC Aug 26 00:26:33 03tmbinc 07org.oe.dreambox * rfc81d4b3... 10/ (1 packages/initscripts/initscripts-opendreambox_1.0.bb): initscripts-opendreambox: install var.tar.gz into /etc Aug 26 00:26:43 03tmbinc 07org.oe.dreambox * rd69a7ad4... 10/ (1 packages/glib-2.0/glib-2.0_2.12.0.bb): glibc-2.0.12.0: fix SRC_URI Aug 26 00:26:48 03tmbinc 07org.oe.dreambox * r0bc758ce... 10/ (1 packages/dreambox/dreambox-dvb-modules.bb): dreambox-dvb-modules: remove not anymore required script from SRC_URI Aug 26 00:26:53 03tmbinc 07org.oe.dreambox * r18ae5dc6... 10/ (1 conf/machine/dm600pvr.conf): dm600pvr: use binutils 2.15.91.0.2 which are available Aug 26 00:26:59 03tmbinc 07org.oe.dreambox * raddb7b83... 10/ (1 packages/enigma/enigma_cvs.bb): enigma: add proper PACKAGES_DYNAMIC Aug 26 00:27:08 03tmbinc 07org.oe.dreambox * rb0f056d6... 10/ (3 files in 3 dirs): initscripts-opendreambox: add dm600pvr support **** ENDING LOGGING AT Sat Aug 26 02:59:57 2006