**** BEGIN LOGGING AT Fri May 12 09:59:56 2006 May 12 10:08:58 someone tried pxa27x cpufreq? May 12 11:06:56 morning May 12 11:07:07 hi ken May 12 11:09:15 XorA: is there a way to check FPS in mplayer? May 12 11:11:51 hrw|work: cant remember May 12 11:12:16 hrw|work: but if its not dropping frames its running at the movies FPS May 12 11:45:24 good day May 12 11:47:50 hi May 12 11:48:55 wb zecke May 12 11:48:57 hi Cockroach- May 12 12:01:45 Hi all May 12 12:01:50 hi dirk May 12 12:01:58 hey marcin May 12 12:25:05 koen: your 'ipaq-pxa270' is hx4700 with 2.6.16-hh? May 12 12:44:11 hrw|work: yes May 12 12:44:40 thx May 12 12:44:55 hrw|work: or hx2xxx is you set linux-oz as kernel May 12 12:44:56 koen: did you had a chance of testing my changeset? May 12 12:45:03 not yet May 12 12:45:11 I'm at my parents now for the weekend May 12 12:45:58 no problem May 12 12:55:52 koen: card is 'asus wl-110' - right? May 12 12:56:13 hi hrw May 12 12:56:16 * hrw|work prepares patchset for netdev list May 12 12:56:17 hi woglinde May 12 12:56:37 hm what to do to enable glibc 2.4? May 12 12:57:01 PREFFERED_VERSION_glinc = "2.4" didnt work May 12 12:57:48 glinc? May 12 12:57:53 glibc May 12 12:57:55 typo May 12 12:57:58 her eon irc May 12 13:00:04 hm next try with the entry from angstroem conf May 12 13:00:11 hi mickeyl May 12 13:00:12 jo mickeyl May 12 13:00:18 hey mickeyl May 12 13:00:24 mickeyl: which prism2 cards you have? May 12 13:00:30 mickeyl: do you still have a binary sdk for opie somewhere? May 12 13:00:50 hi guys May 12 13:01:13 hrw|work: SanDisk 0MB OEM, SanDisk 128MB ConnectPlus May 12 13:01:39 lardman|work: you mean something like what I did for 3.5.3? A copy of a staging area? May 12 13:02:21 mickeyl: Yes, that kind of thing. I think it would be worthwhile to get more people using opie/gpe without needing to learn about oe May 12 13:02:36 mickeyl: people keep asking for toolchains May 12 13:02:53 hrms May 12 13:02:55 NOTE: preferred version 2.4 of glibc not available May 12 13:04:33 lardman|work: *nod* i can imagine. for opie it's tough to do anything else than the staging are bundle, because all our -dev packages suck May 12 13:05:00 mickeyl: so hrw told me May 12 13:05:37 lardman|work: i no longer have anything 3.5.4 here, but hrw could easily tar up his staging area. hrw? May 12 13:06:10 mickeyl whats the diffrent btw. glibc and glibc-intermediate May 12 13:06:27 I can May 12 13:06:28 woglinde: glibc is the real thing, glibc-intermediate is necessary for toolchain generation May 12 13:06:31 but after 3.5.4.1 May 12 13:06:34 hrw|work: cool. thanks. of course May 12 13:06:41 hrw|work: ETA for .1? May 12 13:06:45 hrw|work: No rush, just something that would be useful May 12 13:07:56 hey all May 12 13:08:03 mickeyl: if find time during weekend then next week May 12 13:08:29 k. is there something you want me to test ? May 12 13:10:37 mickeyl: 3.5.4.1-rc2 May 12 13:10:58 which model in particular? May 12 13:11:18 (i can offer c7x0,tosa,spitz,collie) May 12 13:11:22 cxx00 preferred May 12 13:11:27 mickeyl: so spitz May 12 13:11:30 k May 12 13:11:59 * hrw|work exchange collie for h22xx or any qvga ipaq running 2.6 and supported well May 12 13:13:40 can anyone tell me if OE works for x86 targets? like a normal pc ? May 12 13:14:07 webmind: in theory yes. MACHINE = epia is an example. in practice, it's not actively maintained atm. May 12 13:14:26 hmm May 12 13:15:09 sounds like you're our man ! :D May 12 13:15:15 where can i post bug for matchbox ? May 12 13:15:23 is there a forum somewhere ? a list ? May 12 13:15:31 mickeyl, what's the maintanance work for it ? May 12 13:15:57 webmind: no idea, to be honest. checking if stuff like opie-image or gpe-image still compiles for a start May 12 13:16:16 k May 12 13:16:23 might look into that then May 12 13:16:43 it compiles but it does not run :/ May 12 13:16:54 ow :( May 12 13:16:57 the x86-kernel is in need of...repairs May 12 13:17:07 only the kernel ? May 12 13:17:13 hi ifaistos May 12 13:18:11 OE seems like a good alternative for low performance setups May 12 13:18:28 basicaly the kernel but also other stuff need fixing. Udev was not working for example May 12 13:18:45 woglinde : Hi ! May 12 13:18:46 hmk May 12 13:18:59 webmind : It is :) May 12 13:19:45 I'm trying to compile angstrom but osb-jscore can't be downloaded because "/home/juarez/.cvspass for reading: No such file or directory" May 12 13:19:51 any tip? May 12 13:20:47 hm try touch $HOME/juarez/.cvspass May 12 13:22:15 mickeyl: you did the libqpe-opie.bb file right? May 12 13:23:53 Anyone with commit rights that can do a quick fix? May 12 13:24:11 katossi: touch it May 12 13:24:44 KenJlime: yes, libqpe-opie is mine May 12 13:24:52 mickeyl: it lacks depend for qte May 12 13:25:17 see #982 May 12 13:25:27 libqpe-opie.inc:DEPENDS = "uicmoc-native virtual/libqte2" May 12 13:25:31 looks ok to me May 12 13:25:45 It doesnt work for me, check the buildlog May 12 13:25:57 when installing qte before libqpe it works May 12 13:27:21 sounds like you need to set May 12 13:27:39 QTE_VERSION May 12 13:27:50 PALMTOP_USE_MULTITHREADED_QT May 12 13:27:59 the former being a version, the latter being yes or no May 12 13:28:05 then rebuild from scratch May 12 13:28:19 Using familiar_unstable as distro setting currently May 12 13:28:31 familiar-unstable looks unmaintained in .dev May 12 13:29:03 oki, but its a distro bug, not a package bug? May 12 13:29:09 hrw|work: I have no such .cvspass May 12 13:29:25 katossi: touch will create it May 12 13:29:36 katossi hm I was wrong May 12 13:29:38 mickeyl: that case ill just drop the bug May 12 13:29:45 try touch $HOME/.cvspass May 12 13:30:24 KenJlime: yeah. actually it should have a reasonable without these settings, but it seems it doesn't. May 12 13:30:37 i can check what happens when i remove the settings from my distro May 12 13:30:39 * mickeyl rebuild May 12 13:31:21 mickeyl: oki,thx May 12 13:33:59 KenJlime: can you check whether you have libqte-mt in STAGING as well? May 12 13:34:33 STAGING_DIR/sh3-linux/qt2/lib May 12 13:34:36 iirc May 12 13:36:04 got 4 libqte files, but no libqte-mt file May 12 13:36:19 libqte.so, libqte.so.2, libqte.so.2.3, libqte.so.2.3.10 May 12 13:36:32 hmm May 12 13:36:46 let me checkin something and then you could do a May 12 13:36:52 bitbake -nv libqpe-opie May 12 13:36:57 and pastebin the log please May 12 13:37:04 yeah May 12 13:38:13 hrw, woglinde it worked, thx May 12 13:38:34 katossi no prob May 12 13:38:36 np May 12 13:39:09 mickeyl: http://bugs.openembedded.org/attachment.cgi?id=739 May 12 13:39:54 I keep getting the following Linux Kernel error right after the kernel uncompresses: Error: unrecognized/unsupported machine ID (r1 = 0x000003e5). May 12 13:40:03 KenJlime: wait wait... it's not comitted yet :) May 12 13:40:22 ok, it's up May 12 13:40:28 ah lol May 12 13:40:28 I have the correct machine set in the kernel config file and I verified it is in the arch/arm/tools/mach-types May 12 13:40:28 list May 12 13:40:53 I'm not sure why it sees 3e5 as the machine since that is not even a valid machine. May 12 13:40:56 * hrw|work just sent 5 hostap/orinoco patches to netdev ML - time to wait for fscks May 12 13:42:08 mickeyl: want me to post diff against the "old" nv i did, or just complete? May 12 13:42:13 03mickeyl 07org.oe.dev * rb9292e45... 10/conf/distro/preferred-opie-versions.inc: preferred-opie-versions.inc: set overridable defaults for QTE_VERSION, OPIE_VERSION and PALMTOP_USE_MULTITHREADED_QT May 12 13:42:17 03mickeyl 07org.oe.dev * rbb64a8d0... 10/conf/distro/familiar-unstable.conf: familiar-unstable.conf: unbreak a bit May 12 13:43:40 KenJlime: complete is better for me to read :) May 12 13:43:47 KenJlime: which version of qte was built? May 12 13:44:27 mickeyl: oki, uploading shortly May 12 13:44:55 zecke: the problem was that no qte was built for me, and thus libqpe couldnt link against it May 12 13:45:27 uhm, there is another message that lists valid machines and the only one in the list is the one I set in the config file. So does this mean that the error I see is because the kernel gets the actual machine from the bootloader? How does that work? May 12 13:46:00 he zecke May 12 13:46:00 Gerrath: i guess you need to read the section about booting in linux/doc May 12 13:46:17 morning May 12 13:46:18 Gerrath: it's a bit longish to explain here and my fingers are already aching May 12 13:46:22 hi zecke May 12 13:46:40 mickeyl, no problem I would rather read it anyway. This is a doc in the kernel source? May 12 13:46:47 Gerrath: yea May 12 13:46:55 damn, i need a faster machine May 12 13:47:01 especially the arm document is a pretty good read May 12 13:48:22 mickeyl: http://bugs.openembedded.org/attachment.cgi?id=740 May 12 13:48:27 mickeyl, Doc: /documentation/arm/Booting, thanks. May 12 13:49:42 mickeyl: I could clean out libqpe and qte and see if it grabs qte now when building libqpe May 12 13:50:43 woglinde: hey May 12 13:50:46 woglinde: *toothache* May 12 13:51:00 zecke will it ever stop? May 12 13:51:01 mickeyl: Just did and it grabbed qte May 12 13:51:08 KenJlime: ok, good May 12 13:51:30 woglinde: sure May 12 13:51:39 mickeyl: thx, yet another bug squashed. I got 2 dead packages from the goal of having complete opie May 12 13:52:15 zecke I try to build a image with eabi and glib2.4 for the simpad May 12 13:53:37 woglinde: harald is a idiot, putting the muxer into the kernel... May 12 13:53:47 an May 12 13:53:51 by the way May 12 13:53:52 right ;) May 12 13:53:58 hm which muxer May 12 13:54:05 I rad the blog May 12 13:54:07 read May 12 13:54:18 woglinde: GSM/ETSI May 12 13:55:25 KenJlime: cool. progressing nicely May 12 13:55:29 hm oh i missed some entries May 12 13:55:36 zecke: which harald? May 12 13:55:40 welte May 12 13:56:31 ah. then i had the right one in mind May 12 13:56:47 he seems to be a pretty competent guy in general... but slightly arrogant as well :) May 12 13:56:57 unfortunately this is a usual combination May 12 13:57:12 s/usual/not a rare/ May 12 13:57:14 RMS May 12 13:57:15 :D May 12 13:57:27 sounds like a kernel hacker disease May 12 13:57:31 RP being one notable exception May 12 13:57:32 ~praise RP May 12 13:57:34 All hail RP! May 12 13:57:37 hihi May 12 13:57:41 lol May 12 13:59:03 03mickeyl 07org.oe.dev * r953b27f3... 10/conf/ (3 files in 3 dirs): May 12 13:59:03 add distro configuration for EZX phone distribution May 12 13:59:03 eventually this will be merged with Angstrom May 12 13:59:40 hms May 12 13:59:44 damnit glibc-build May 12 13:59:47 mickeyl: I added .bb file for blackbox, its not pretty but compiles. Should be good for a fixme start May 12 13:59:58 bits/fenv.h not found May 12 14:01:31 harald is the one behind openezx? May 12 14:01:37 hrw yes May 12 14:02:56 hey mickeyl May 12 14:04:14 hrw|work: yes, wl-110 May 12 14:04:29 koen: ok thx May 12 14:04:34 KenJlime: good to hear May 12 14:04:35 koen: morning May 12 14:04:40 * koen searches for sunscreen May 12 14:04:54 hi koen May 12 14:04:59 mickeyl: its up on the buglist, you can implement it when you got time May 12 14:05:03 koen: http://marc.theaimsgroup.com/?l=linux-netdev&m=114744109711281&w=2 May 12 14:05:05 KenJlime: will do, thanks May 12 14:05:10 * mickeyl wanders home May 12 14:05:13 bbl May 12 14:05:20 cu mickeyl May 12 14:06:13 * koen reads May 12 14:06:16 hey KenJlime May 12 14:06:33 mickeyl: well, I started to wonder about the competence :} May 12 14:10:54 Anyone else having build segfaults with opie-rotateapplet? May 12 14:11:11 mickeyl: *hop* *hop* *hop* ;) May 12 14:11:16 KenJlime: gcc segfaults? May 12 14:11:58 bug #973 May 12 14:12:20 zecke: will it work if I will add extra field into SRC_URI line? "file://04-netgear-ma701.patch;patch=1;comment=sent-to-netdev" May 12 14:13:21 KenJlime: Why do you use 1.5.0? ;) May 12 14:13:34 hrw|work: it should, just avoid spaces May 12 14:13:42 ok May 12 14:14:07 zecke: I like bleeding edge :D Its worked well. May 12 14:14:21 KenJlime: and it might stop working any time May 12 14:14:38 KenJlime: okay, gcc.gnu.org and see what you need to submit gcc errors May 12 14:14:46 KenJlime: e.g. get the parse treee etc.. May 12 14:14:55 zecke: roger, thx May 12 14:15:48 ~fsck sync on sd card May 12 14:15:50 e2fsck /dev/sync on sd card : warning! filesystem contains idiots! May 12 14:16:52 haha May 12 14:21:25 best infobit ever May 12 14:29:26 where is DN_NOTIFY,... and such defined? May 12 14:44:19 woglinde: ping May 12 14:49:07 hmm... we should set up vim-tiny to replace bb vi if it exists May 12 14:52:09 how small is vim-tiny? May 12 14:53:30 debian vim-tiny/arm = 906K May 12 14:53:39 whole busybox is <600K iirc May 12 14:53:56 yeah, yeah - I know May 12 14:54:11 but I've been getting death threats lately over the busybox vim May 12 14:54:34 heh May 12 14:54:44 busybox vi is a bit ..... May 12 14:54:46 @^%$&@^$&*@^$&# May 12 14:54:58 I couldn't have said it better.. May 12 14:55:18 I like this vi May 12 14:55:32 our OE stuff is mature enough that its starting to roll out to the less - Linux friendly among us May 12 14:56:11 * koen wonders how busybox emacs behaves May 12 14:56:16 one great fix for busybox vi is to remove the hardcoded TERM=vt100 May 12 14:57:40 valuable tip May 12 14:58:06 CosmicPenguin: look in /etc/profile May 12 14:58:43 CosmicPenguin: that upgrades it from totally unusable to vaguely usable May 12 14:59:03 wow - thats way better May 12 14:59:09 good to know May 12 14:59:48 Does any of you actually use emacs? May 12 15:00:01 no May 12 15:00:13 I used to use emacs to flush my buffers May 12 15:00:20 gnu emacs? hell no May 12 15:00:25 start emacs, quit emacs -> 32MB freed May 12 15:00:26 * pb__ xemacs May 12 15:00:31 * XorA used to use microemacs on amiga May 12 15:00:44 The times ive used it I always get lost in the ctrl-alt-p-q-t-f-w-r = take the dog out for a walk syntaxes May 12 15:01:45 XorA: I used vim on amiga May 12 15:01:56 XorA: sometimes - mostly golded or cygnused May 12 15:02:10 hrw|work: I dont think anyone had ported wim back in the 80's May 12 15:02:39 * CosmicPenguin mumbles something about writing code in emacs May 12 15:02:40 hrw|work: though I didnt have an internet connection to get it anyway :-) May 12 15:02:48 but I use VI for most everything else May 12 15:02:48 :) May 12 15:04:08 XorA: vim was created on Amiga May 12 15:04:38 XorA: then ported to other OS May 12 15:04:56 Ciao all May 12 15:05:02 hi Pigi May 12 15:05:24 hi hrw|work May 12 15:05:45 XorA: 1991 Nov 2 Vim 1.14 First release (on Fred Fish disk #591). May 12 15:06:41 hrw|work: cool May 12 15:06:42 XorA: 1988 1.0 Based on Stevie for the Amiga, named Vi IMitation, never publicly released May 12 15:06:52 XorA: http://en.wikipedia.org/wiki/Vim_(text_editor) May 12 15:08:09 hrw|work: my dissertation was done in Vim on Amiga in 97 May 12 15:09:08 so, has anyone ever thought about messing with GCC spec files in order to be able to buld uclibc and glibc in the same environment and switch between them? May 12 15:09:13 or is that just a horrible, scary evil idea? May 12 15:09:49 CosmicPenguin: it's a nice idea May 12 15:09:50 Pigi: does ipkg .162 worth upgrading from .159? May 12 15:10:00 CosmicPenguin: good thing you volunteered May 12 15:10:19 hrw|work, there are a lot of interesting bugfix ..... May 12 15:10:30 koen: I'm thinking about it - we're starting to build initrds, and these things pop into my mind... :) May 12 15:10:45 CosmicPenguin: not sure about horrible, scary or evil, but it may be a futile idea. I'm not sure that all the differences between those two environments are encapsulated in the specs file. May 12 15:10:58 sad May 12 15:10:59 you're probably better off just parallel-installing two versions of gcc, one for each May 12 15:11:08 yeah - that might not be that bad too May 12 15:11:16 Pigi: did you see http://bugs.openembedded.org/show_bug.cgi?id=857 ? May 12 15:11:33 My more pressing concern is getting a simple version of ipkg that doesn't write all the control data into the intrd May 12 15:11:48 koen, lemme see May 12 15:11:49 stupid x86 - always a pain in the hey-hey! May 12 15:11:51 can you not just post-process the initrd to delete all that data? May 12 15:12:09 pb_: you sound like my boss, what with the smart ideas and everything... :) May 12 15:12:15 Let me have my dreams! May 12 15:12:35 heh May 12 15:13:21 koen, I thinks some ( if not all ) of this patches are in .162 May 12 15:13:38 CosmicPenguin: try teaching bb to make libbb shared, so ipkg doesn't have to include its own copy May 12 15:14:30 Pigi: http://oe.pastebin.com/713618 - 1.5m to get info that package does not exist ;( May 12 15:14:50 * hrw|work want faster nand.... May 12 15:15:26 I think I can't speedup things if package file is big. May 12 15:15:44 Pigi: ~6k packages May 12 15:15:45 In this case, I do need to read trough it to see if the package is there. May 12 15:16:30 hrw|work, 192k packakges May 12 15:16:41 6k lines May 12 15:17:01 Pigi: no - 6k lines of "Package: name" so 6k of packages May 12 15:17:07 192k chars May 12 15:17:09 yes, I've seen May 12 15:17:29 hi May 12 15:17:41 btw 6k packages isn't a little amount of data for it. How big is the file ? May 12 15:18:50 Pigi: 2.389.170 bytes in total (few feeds summarized) May 12 15:20:20 I don't really know how to speed up things. How much does it takes to grep this file ? May 12 15:21:07 6s May 12 15:22:29 hrw|work, as you probably knows, whenever a packages file get read, lots of time is spent in creating the needed ( nested ) structure to support the install May 12 15:22:44 Pigi: I know May 12 15:23:45 Pigi: no ideas for fixing May 12 15:23:51 * hrw|work need to go May 12 15:23:53 cu all May 12 15:24:00 thus, when reading trough 6k Packages, it must allocate a very hyuge amount of memory May 12 15:24:07 ciao hrw|gone May 12 15:25:33 Pigi: ever considered storing the lists in a sqlite db? May 12 15:27:41 koen, I'm not sure it will change something. The problem is not only in reading through the file, but in allocating memory for the variuos struct. May 12 15:28:56 Pigi: have you considered using a more efficient allocator for that? say, GSlice or some such May 12 15:29:33 pb__, honestly not. But I could try to read about this. May 12 15:30:07 http://developer.gnome.org/doc/API/2.0/glib/glib-Memory-Slices.html May 12 15:35:09 a+ May 12 15:42:19 re zecke May 12 15:44:20 spaciva May 12 15:44:27 woglinde: genosse May 12 15:45:00 hmm what can we do (without using launchpad) to make bug reports go to the right place (mmagically) May 12 15:46:33 bbl May 12 15:53:23 Well lost my primary connection, but his is me off for 2 weeks May 12 15:53:27 see you dudes later! May 12 15:53:57 cya! May 12 16:01:43 is there a problem with the forums? May 12 16:01:51 which one? May 12 16:02:01 oesf May 12 16:02:04 *g* May 12 16:02:06 the whole site May 12 16:03:03 YoG: dunno May 12 16:03:42 well... its working really slow now (for me at least) May 12 16:05:22 pb_: master of gnu? May 12 16:06:16 zecke: hail zecke May 12 16:06:45 pb__: could you do me a favor and open gcc/c-incpath.c May 12 16:07:28 sure May 12 16:07:31 which version of gcc? May 12 16:07:36 pb__: any May 12 16:07:50 okay, got it May 12 16:07:59 pb__: I could add a hack to add_path to check if the string starts with "/usr/" May 12 16:08:00 right? May 12 16:08:07 Could we drop java support in the gcc packets (intended for target arch). They are giving me a world of trouble May 12 16:08:14 does this sound like a good idea? May 12 16:08:20 zecke: yes, sounds like a good plan May 12 16:08:27 KenJlime: you can drop it for sh May 12 16:08:33 KenJlime: you can turn them off for superh if you want May 12 16:08:46 KenJlime: gcc*3*.inc May 12 16:08:49 any attempt to turn them off for other architectures would be met with fierce resistance May 12 16:09:38 how can I do that? May 12 16:11:44 KenJlime: gcc3-build.inc May 12 16:11:48 ~line 14 May 12 16:12:11 and for gcc4 as well May 12 16:12:13 food May 12 16:12:18 will hack gcc later May 12 16:12:19 oki, thx May 12 16:12:32 and I need to hurry otherwise Al will get mad May 12 16:39:42 pb_ it compiles nicely, was due to nasty shit libgcj... just left that out as for most archs May 12 16:40:09 I've sent a patch if anyone care to commit it (Bug #977) May 12 16:55:55 hi May 12 16:56:32 I have to decide one thing about wifi cards in 2.6... follow mainline or be against it. May 12 16:56:50 how is the mainline different from what you want? May 12 16:56:52 what you mean with mainline? May 12 16:57:03 woglinde: 2.6 May 12 16:57:25 raduga: prism2 cards can be used with hostap or orinoco May 12 16:57:53 hrw hm thats the problem waiting until all drivers are ported or made some other May 12 16:57:54 mainline supports just one or the other? or mainline supports both, and you want to specify one or the other? May 12 16:58:04 raduga: some are handled by both - with pcmciautils we have them both loaded and some cards goes to orinoco instead of hostap - so no WPA May 12 16:58:18 anyway I got one nice suggestion - blacklisting May 12 17:02:06 how does that blacklisting work? May 12 17:02:31 the whole 'can only change stuff in kernel' approach seems a bit odd May 12 17:02:43 koen hm eabi cannt be build for arm4v right? May 12 17:03:01 woglinde: in OE it can, the thumb bits aren't turned on May 12 17:03:17 woglinde: see http://wiki.debian.org/ArmEabiPort for details May 12 17:03:37 koen but glibc says it cant use the compiler May 12 17:03:46 weird May 12 17:04:05 it works for my armv4t board ;) May 12 17:04:15 hm May 12 17:04:19 with angstroem? May 12 17:04:29 yes May 12 17:04:53 koen: hmm. just checked and it looks like blacklisting in hotplug is not that nice May 12 17:05:05 its rather 'do not use this driver at all' May 12 17:05:12 koen can you provide me the local.conf? May 12 17:05:13 hrw: that's what I thought May 12 17:05:35 woglinde: DISTRO="angstrom-2006.9" MACHINE="yourmachine" May 12 17:05:39 nothing more should be needed May 12 17:06:09 koen trying I am trying for simpad May 12 17:06:24 ah May 12 17:06:32 the 2.6 kernel for simpad doesn't build May 12 17:06:39 so you're out of luck anyway May 12 17:06:51 koen the kernel builds May 12 17:07:00 but it only works for zecke May 12 17:07:16 because of the fb driver May 12 17:09:53 i have a question, regarding an annoyance in wifi; May 12 17:10:03 not sure to whom to address, or how to address May 12 17:10:03 koen: I "fixed" the glibc-intermediate problem I stumbled on earlier for my case. May 12 17:10:15 http://www.oesources.org/source/current/libXcomposite-X11R7.0-X11R7.0-0.2.2.2.tar.bz2 can't be downloaded, can anyone try if it's just me? May 12 17:10:21 i'd like to blacklist some "bad" access points, and some bad networks May 12 17:10:45 my card keeps randonly switching APs, and some of the ones it locks onto, are useless, and won't serve me at all May 12 17:11:08 and there are a few, specific ones that I'd like to use, but generally, any other than the "bad" APs are fine May 12 17:11:25 i can't seem to find any way in drivers/software to enforce that kind of policy May 12 17:11:51 raduga I think you are right May 12 17:11:51 katossi: it is a noted bug. May 12 17:12:00 ok May 12 17:12:20 katossi: fixed, in .dev May 12 17:12:39 katossi: you can either, pull newer .dev from monotone, May 12 17:12:40 I'm using .dev May 12 17:12:46 hi florian May 12 17:12:47 or patch your .bb manually May 12 17:12:51 I will pull now May 12 17:12:51 whichever you prefer May 12 17:12:55 :) May 12 17:14:04 http://bugs.openembedded.org/show_bug.cgi?id=965 May 12 17:15:28 re May 12 17:18:12 koen ah now I understand the kernel needs eabi too May 12 17:25:03 like|food4me: iirc I patched the patch: http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=1ad17d00731f7914d50ac92436fc537276cf1f9c May 12 17:36:28 NOTE: package quilt-native-0.45-r0: task do_build: completed May 12 17:37:13 koen: gracias May 12 17:38:04 gah May 12 17:38:10 ERROR: log data follows (/data/build/koen/OE/build/tmp/angstrom/work/armv5te-linux-gnueabi/bzip2-1.0.2-r1/temp/log.do_patchcmd.27982) May 12 17:38:13 | getopt: invalid option -- n May 12 17:38:15 | Usage: quilt import [-p num] [-P patch] [-f] [-d {o|a|n}] patchfile ... May 12 17:38:36 hm.. May 12 17:42:05 comment added to #929 May 12 17:44:18 koen: so I guess it wasn't a problem me pushing to oz354x? May 12 17:46:05 JustinP: don't ask me :) May 12 17:46:50 koen: heh, ok May 12 17:50:21 btw, I built a gpe-image-akita yesterday, with src a couple days old, May 12 17:50:54 and the kernel had *major* problems in dealing with the jffs2 at mount time. Panicked, and everything. May 12 17:51:24 interesting May 12 17:51:28 not reporting it to the bugtracker, as I'm not positive it was a kernel problem (yet), or some sort of pilot error. May 12 17:51:35 monotone: 1,8 M | 232,2 k | 1684 | 407 | 24 May 12 17:51:37 ;( May 12 17:51:47 hrw: it was just after trying sashz' beta4, so I am a little skeptical May 12 17:52:10 raduga: u-boot on device or reflashed whole nand? May 12 17:52:21 hrw: beta4+uboot -> nand flash with Sharp rom -> oz .dev May 12 17:52:29 hrw: the latter May 12 17:52:31 ok May 12 17:52:55 hrw: there's also a possibility of bad CF card. too many variables May 12 17:53:12 sashz' beta4? what's this? May 12 17:53:12 i am buying a brand-new CF, and building a brand-new gpe, May 12 17:53:34 so we can eliminate some of the variables. i'll send report to bugs list, if it persists :) May 12 17:53:49 JustinP: pdaXrom beta4 May 12 17:53:57 justinp: with u-boot. May 12 17:54:48 * koen looks at the weather forecast May 12 17:55:02 it seems I'll have some time on sunday to play with qemu May 12 17:58:19 KenJlime: #984 will stay for long time in bugtracker probably May 12 18:04:44 hrw: howso? May 12 18:06:17 KenJlime: until next bugparty when someone will take it as 'simple bug which would be nice to resolve and I have time to check which ver is current' May 12 18:06:54 KenJlime: in other words - if package is unmaintained and not popular it can stay for long time (if there is no patch to fix bug attached) May 12 18:10:10 KenJlime: btw - what do you think about having r/w access? May 12 18:10:19 hrw: oki, which basicly means that I will look into then :D May 12 18:11:26 koen: if/when you got time got #977 and #980 ready for commiting May 12 18:11:47 koen: thanks for the arm-longlong patch fix. pulling. May 12 18:13:38 mickey|bbiab: having same issues with xmms-embedded as I had with libqpe-opie (lacking -lqte) could you do your quick magic fingers on that one also? May 12 18:15:04 KenJlime: vi xmms-embedded*.bb and remove 'CONFIG-=thread" from EXTRA_QMAKEVARS_POST and do rebuild May 12 18:15:15 KenJlime: tell does it helps May 12 18:16:53 hrw: didnt make any difference May 12 18:17:06 check log to see what mickey did for libqpe-opie May 12 18:17:09 had exact same issues May 12 18:17:21 03hrw 07org.oe.oz354x * r05815600... 10/packages/totem/totem_1.0.4.bb: totem: disable lirc - we have it only on collie May 12 18:19:16 going to install qte and then xmms-embedded, fixed it with libqpe-opie although mickey will need to fix the xmms-embedded..bb file later May 12 18:35:18 raduga: and u-boot is.....? May 12 18:38:07 JustinP: bootloader May 12 18:42:54 Need your ideas.... Once the images for my distro are build i end up with 2 (actually more) files one for the kernel and one for fs. In order to create a binary image a user can dd to a CF or a disk, i need to create 2 partitions in a CF, format them,run grup, install the kernel in the first partition, install the fs in the second and them run dd to create a binary file with an image of the compact flash. May 12 18:43:13 is there a way to automate the procedure from OE ? May 12 18:43:29 probably May 12 18:43:32 oe do i need to extend the makefile i have May 12 18:44:00 oe=or May 12 18:44:03 Ifaistos: you want to create disk image with two partitions May 12 18:44:20 hrw: yes May 12 18:44:21 iirc we had something for this added some time ago May 12 18:45:20 hrw : in which module/package ? May 12 18:46:27 do not remember now ;( May 12 18:48:49 hrw : let me know if something comes up.... May 12 18:48:56 On another matter.... May 12 18:49:22 I am building the same distro for several different machines and architectures... May 12 18:49:36 each time i have to go through building the tool-chain May 12 18:50:09 is there a way to have a common tool-chain for all and generate different packages etc May 12 18:50:43 multimachine.inc May 12 18:52:06 hrw : maybe multimachine.conf ? May 12 18:52:27 perhaps - I have it inside of local.conf May 12 18:54:20 hrw : Can you post it please ? May 12 18:58:03 http://oe.pastebin.com/714032 May 12 18:59:28 hrw : Thanks :) May 12 19:13:10 all: Anyone compiling (Angstrom) from a x86_64 host? May 12 19:13:49 I would like ;D May 12 19:14:29 hrw: I could let you login :-) May 12 19:15:17 what problem you got? May 12 19:15:25 building gm-native gives me: configure: error: some sort of compiler problem, mp_limb_t doesn't seem to work May 12 19:15:53 <6>note: kio_http[11373] exited with preempt_count 3 May 12 19:15:59 KILL kde May 12 19:16:00 hrw: which is preceded by: checking for mp_limb_t... yes | checking size of mp_limb_t... 0 May 12 19:16:27 so obviously the size 0 for a limb is not good. May 12 19:16:41 then check mp_limb_t and set proper var in site/ file? May 12 19:18:50 hrw: thanks for the suggestion - my build skills are limited May 12 19:19:03 happens May 12 19:22:47 * JustinP waits for it May 12 19:23:51 03justinp 07org.oe.dev * r42b8c792... 10/ (64 files in 7 dirs): efl, e17: update to newest versions May 12 19:24:10 ^_^ May 12 19:24:23 * JustinP has this update thing scripted now ;-) May 12 19:25:10 http://www.oesources.org/source/current/JavaScriptCore_cvs.sourceforge.net__20060212.tar.gz can't be downloaded May 12 19:25:35 katossi: the stuff from oesources is just a bonus May 12 19:25:49 katossi: bb/oe will also check out from CVS... May 12 19:25:57 (assuming sf.net fixed their servers) May 12 19:27:11 but it's not a bug May 12 19:28:18 did it download from CVS ok? May 12 19:29:36 no May 12 19:30:00 cvs [checkout aborted]: end of file from server (consult above messages if any) May 12 19:30:00 well.... May 12 19:30:07 wonderful sf.net CVS May 12 19:30:56 cvs.sf.net is changing slowly May 12 19:31:08 wt.....ERROR: Nothing provides dependency libx11 May 12 19:31:26 JustinP: x11<>libx11 diff .dev<>.oz354x May 12 19:31:35 ah May 12 19:31:38 ok, then May 12 19:31:57 katossi: OZ 3.5.4 has source archive, familiar 0.8.x too May 12 19:31:59 I also think this should be virtual/x11 as (AFAIK) it's impossible to build non-diet x11 without using -b May 12 19:34:56 bbl - have to reboot... May 12 19:35:37 all: how can I tell oe / bitbake to compile gmp-native 4.2.1 for me instead of the angstrom default 4.1.2? May 12 19:36:06 all: i.e. where can I find the version number being built? May 12 19:36:14 PREFERRED_VERSION_gmp-native = "4.2.1" in local.conf I think May 12 19:37:43 JustinP: thanks. But where is the default version taken from? (I am quite newbie here) May 12 19:42:32 hi May 12 19:42:38 hrw|gone: you're around ? May 12 19:48:22 hi again May 12 19:58:47 is it possible to run something like "bitbake gpe-image --download-all-sources" to just grab all sources? May 12 20:02:53 sure May 12 20:03:01 bitbake -c fetch or so May 12 20:03:05 take a look at bitbake -h May 12 20:17:04 03mickeyl 07org.oe.dev * rb4da6e3e... 10/packages/ltrace/ltrace_0.3.36.bb: ltrace: remove do_install May 12 20:17:09 03mickeyl 07org.oe.dev * r8c29439f... 10/conf/ (3 files in 3 dirs): tweak ezx configs. bootstrap-image now builds, further tweaks not until I got my A780 :) May 12 20:17:51 mickey|tv: somehow I doon't believe the 'no further updates' part May 12 20:18:00 hehe May 12 20:18:01 mickey|tv: and where did you buy the a780? May 12 20:18:09 I can't find a retailer in .nl May 12 20:18:18 * koen also need a new phone May 12 20:18:26 ordered @ hitseller.de May 12 20:18:55 http://www.hitseller.de/shop.php?mode=show_detail&group=9D2DB2ADE4D3187C95E963B1D109AC7E&last_group=C616BB7E97877F068557B6D8393ADC7A&sid=4e41ef739766cc309e66ccc0aef4256f&s=&id=1D2EC2B723A0A84287191B8E7269D203 May 12 20:19:12 they say it's in stock May 12 20:19:12 eeks! May 12 20:19:20 ordered it on wednesday May 12 20:19:25 can't wait to have it here May 12 20:19:27 that's like a week worth of pay May 12 20:19:28 * mickey|tv gearlust 100% May 12 20:19:39 yay, pretty expensive :/ May 12 20:19:50 need that to improve my mood ;) May 12 20:19:52 the blueangel is E650 May 12 20:22:11 harald welte making good progress w/ ezx May 12 20:22:20 kernel 2.6 already boots and runs a bootstrap from transflash May 12 20:22:43 without touching your installed system May 12 20:22:56 we also have the modified blob for the a780 May 12 20:24:01 KenJlime: yeah, xmms-embedded needs more qte-mt fixing. as do a lot of non-opie programs which i didn't fix yet. i'll take care about that May 12 20:27:56 mickey|tv: it'd be interesting to experiment with vnc/xcb/freenx over gprs :) May 12 20:27:57 mickey|tv: thx, but I saw that you added the blackbox.bb but havent seen it when i updated. May 12 20:28:45 Oh and if you got a sec later got #977 and #980 ready for commiting May 12 20:28:47 koen: yeah. i wish it had wlan as well May 12 20:29:39 mickey|tv: is it ok to upgrade python-pysqlite2 to 2.2.2? May 12 20:29:52 cbrake: sure. please go ahead May 12 20:31:23 mickey|tv: nice work w/ all the python stuff in OE -- really slick :-) May 12 20:31:53 cbrake: thanks :) put much blood sweat and tears into it. May 12 20:33:32 mickey|tv: maybe we should organize a OE-dev phone-doom challenge May 12 20:33:43 mickey|tv: after we all get sponsored phones ;) May 12 20:33:48 koen: heh. good thinking! May 12 20:34:13 hey TheMasterMind1 May 12 20:35:34 afternoon May 12 20:36:25 03mickeyl 07org.oe.dev * rc3fb2b0c... 10/packages/gcc/gcc3-build.inc: gcc3-build.inc: disable JAVA for sh3. fixes #977 May 12 20:36:31 03mickeyl 07org.oe.dev * rc3c78fb6... 10/site/sh3-linux: sh3-linux: add entry for libxfce4util. fixes #980 May 12 20:36:38 mickey|tv: what does "ml" in PR="ml0" mean? May 12 20:36:46 mickey lauer May 12 20:36:55 cbrake: that's the revision for packages which I really really care about May 12 20:37:00 Michael Lauer is my full name May 12 20:37:08 ah, close enough May 12 20:37:14 right :) May 12 20:37:22 mickey|tv: ahh, ok. May 12 20:37:28 likewise: conf/distro probably May 12 20:38:22 JustinP: good work w/ E! any improvements visible? May 12 20:38:54 mickey|tv: ummm....not that I've seen so far....but I haven't checked the changelogs May 12 20:39:08 mickey|tv: also, the default image in .dev will hang X on xtscal May 12 20:39:17 mickey|tv: trying to build a new oz354x image right now May 12 20:39:23 JustinP: tnx May 12 20:39:30 hmmmm..where did login-manager go? May 12 20:39:32 xtscal is broken in .dev due to voodoo toolchain issues May 12 20:39:39 there's a guy @ openezx who is trying to push EFL for a new phone GUI May 12 20:39:45 Is there any reason why there isnt a glibc-2.3.6 package around? May 12 20:39:46 JustinP: it was reverted by 'some people' May 12 20:39:57 dunno whether we really want X on a phone, but we need to check that May 12 20:40:04 mickey|tv: raster already made a phone gui, but for a company May 12 20:40:09 now GET MY PHONE DELIVERED May 12 20:40:13 :D May 12 20:40:25 koen: yes, but that was fb, not x. May 12 20:40:31 mickey|tv: http://www.fluffyspider.com/products/fp/fp.html May 12 20:40:34 * mickey|tv checks display resolution again May 12 20:40:54 heh that's nice May 12 20:40:56 has anyone got a OE port to this yet? http://www.hitseller.de/shop.php?mode=show_detail&lang=de&group=65A75A5CDC19ACC0D130B25AFDED63C2&id=B95C87C68478A8BF307C062E5CCE6C66&last_group=0&sid=4e41ef739766cc309e66ccc0aef4256f May 12 20:41:11 hehehe May 12 20:41:17 likewise: http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=953b27f3fde0c6c63190674f1232e712c0f58d35 May 12 20:41:35 * mickey|tv chuckles May 12 20:41:41 koen: his link points to a beer machine May 12 20:41:52 D'oh May 12 20:41:59 koen: I don't see any commit e-mail about it...:-| May 12 20:42:12 JustinP: happened a few months ago May 12 20:42:14 likewise: when it comes with Leffe Bruin, then I'll add OE support May 12 20:42:16 sorry, it has no display. never mind May 12 20:42:18 bbl May 12 20:42:35 * JustinP wonders why his build worked earlier.... May 12 20:42:43 perhaps I still have an ipk kicking around May 12 20:42:57 mickey|tv: does that offer hold for other platforms as well? We are looking for a build engineer :-) May 12 20:43:06 hmmm.... May 12 20:43:07 nope May 12 20:43:15 ah, I think I re-compiled.... May 12 20:43:20 grrr May 12 20:43:27 so e wouldn't build for anyone anyway May 12 20:43:41 likewise: hehe. actually I'm planning to do freelance work as OE specialist once I submitted my Ph.D thesis May 12 20:44:02 Is there a tool similar to xev for opie so that I can get key names ? May 12 20:44:05 * JustinP now wonders how to fix this May 12 20:44:29 AlexTe: i think so May 12 20:44:32 keyview or so May 12 20:44:36 dunno whether it's in OE yet May 12 20:44:41 because it's a devel tool May 12 20:44:47 i can check tomorrow May 12 20:44:50 seriously, once OE is a bit established (and this goes for other build stuff as well) this is very interesting for embedded companies. May 12 20:44:58 whee. It almost works May 12 20:45:17 JustinP: http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=2e402ac20b6f22afb065088192f491e1bd1b0c09 May 12 20:45:19 likewise: absolutely, that's what I was thinking as well. May 12 20:45:29 k, wifey calls May 12 20:45:32 cu tomorrow May 12 20:45:37 CU May 12 20:45:51 mickey|tv: cu May 12 20:45:58 likewise: agreed -- OE is very interesting for embedded. May 12 20:46:00 (what must his wife have been calling?) May 12 20:46:06 :-) May 12 20:46:09 I would hope so - since its designed for embedded images May 12 20:46:18 sure as hell not interesting for a server farm May 12 20:46:30 CosmicPenguin: Hmm, but you would need one :-) May 12 20:46:42 mickey|tv does not seem to be in oe tree May 12 20:46:45 OE is too bleeding edge to sell to my dev. team May 12 20:46:56 we've been using it for well over a year now May 12 20:47:27 CosmicPenguin: for production software? 24/7 May 12 20:47:35 likewise: what type of products do you work on? May 12 20:47:46 video broadcast equipment May 12 20:47:54 likewise: now we've got experience with branches, a stable branch is getting more real every day May 12 20:48:09 likewise: did you ever do a presentation at fosdem? May 12 20:48:30 as far as I am concerned, the bitbake engine and OE core is stable as can be May 12 20:48:41 koen: it's a bit like with the Linux kernel: new features, new bugs. You need a stable-branch manager. May 12 20:48:50 and if you're doing production images and you're pulling blindly from somebody else's meta data - then you're probably not going to be very successful May 12 20:48:59 koen: no, not yet May 12 20:49:11 likewise: I just started a project for a customer using OE -- small device server type system. May 12 20:49:21 likewise: ok, I remember someone doing a presentation about video stuff in fosdem 2005 May 12 20:50:03 CosmicPenguin: I know. We are seriously looking for someone to maintain our own tree for future development. May 12 20:50:27 likewise: nice to finally have a customer who is interested in using a real build system. May 12 20:51:02 cbrake: nice. Well, I am not that familiar with OE yet. I am trying a handful of environments to come up with one in the end. May 12 20:52:03 likewise: nod, I was referring to the project I mentioned previously. May 12 20:52:10 likewise: good luck with your search. May 12 20:52:19 cbrake: tnx May 12 20:52:26 likewise, cbrake: http://survey.openembedded.org/index/php?sid=6 May 12 20:52:27 03mickeyl 07org.oe.dev * rf74d80c1... 10/packages/xmms-embedded/xmms-embedded_20040327.bb: xmms-embedded: fix SRC_URI per recent sourceforge cvs restructuration, catch up with qte/qte-mt situation May 12 20:52:54 koen: I was meaning in oz354x....but I guess it was never there May 12 20:53:11 koen: which is baffling me because my previous push should have had login-manager and it seemed to build and run fine for me.... May 12 20:53:57 koen: yeah, I already filled it out. Great idea, be interesting to see the results. May 12 20:54:05 mickey|tv: BTW, I now have 2 scripts to automate e updates for me May 12 20:54:18 mickey|tv: one checks the website and spits out mtn rename commands for new versions May 12 20:54:45 mickey|tv: another helps me move changes to branches (oz354x) ^_^ May 12 20:54:56 koen: OK, I'll fill in that form. That aside, what I see currently lacking is documentation on the way you guys handle working with the OE tree. I.e. if something is broken, I am searching through a haystack to see how everything fits together. May 12 20:55:02 cbrake: by the looks of it we don't need bigshot statistics apps to parse the results May 12 20:55:18 cbrake: but we'll let everybody know in a few weeks :) May 12 20:55:54 * JustinP throws out a random "I love OE!" May 12 20:55:58 ~praise OE May 12 20:56:00 All hail OE! May 12 20:56:13 do we have an OE song yet? May 12 20:58:03 koen: heh, its taking awhile, but OE will catch on. May 12 20:58:36 koen: just in the past couple months, I've talked with at least 3 embedded teams seriously considering using OE. May 12 21:04:04 likewise: I think an example step by step debugging procedure would be helpful. May 12 21:04:18 cbrake: yes, exactly. May 12 21:04:33 cbrake: *just* wrote that in similar words in the survey. May 12 21:04:58 You guys seem to be very handy in OE and bitbake already, to me its not that transparent yet. May 12 21:05:07 likewise: OK, I'll try to put something on the wiki in the next few days. May 12 21:05:12 s/its/it's May 12 21:05:44 likewise: I can handle OE/bitbake -- its still autotools that drive me crazy :-\ May 12 21:06:24 cbrake: yes, same here. I am still at the 'make' level of build stuff, there is lots for me ahead yet. And OE builds on all that :-) May 12 21:07:05 cbrake: My biggest problem with the traditional way of cross compiling come along when pkg-config and libtool did no longer work for cross-stuff. May 12 21:07:11 s/come/came May 12 21:07:32 (autoconf and sed :-) syntax) May 12 21:08:15 likewise: nod May 12 21:09:15 from the surbey: biggest problem with OE? => running out of disk space too quicky ;-) May 12 21:09:48 haha May 12 21:10:05 * cbrake thinks we need to combine a bunch of obscure wiki pages into simple FAQ -- opinions? May 12 21:10:31 likewise: heh, reply: disk space is cheap :-) May 12 21:11:14 cbrake: I know, but disk space is never enough :-) May 12 21:12:35 cbrake: a FAQ (or several FAQs) would be nice May 12 21:13:27 cbrake: and more stuff like http://www.angstrom-distribution.org/node/12 May 12 21:14:33 koen: looks good. May 12 21:16:01 * cbrake subscribes to angstrom feed May 12 21:27:47 On the subject of disk space, there is no real harm cleaning up build/tmp/work/ behind bitbake, during the build, right? May 12 21:28:34 likewise: shouldn't be, anything needed by builds *should* be in the staging dir May 12 21:28:41 I mean, the work/* directories are never ever referenced anymore after the package was installed in the staging dir. May 12 21:28:42 (IIRC) May 12 21:29:06 there may be some that still reference work, though, but they should eb fixed May 12 21:29:44 JustinP: ok, I was not sure I might assume that. May 12 21:30:30 Can I override something so that that bitbake does a rm -rf of each working dir after a package was installed to staging? May 12 21:30:52 likewise: There's a class that does that if I remember rightly May 12 21:31:02 likewise: we have some problems with libtool pointing to the workdir May 12 21:31:06 rm_work iirc May 12 21:31:19 rm_work.bbclass to be exact May 12 21:31:31 koen: ok, packages using libtool *might* refer to the work dir still? May 12 21:31:33 s/have/had/ May 12 21:31:52 s/refer/refered/ :-) May 12 21:31:54 likewise: yes, but we should have fixed those issues last month with some sed magic May 12 21:32:25 So using the rm_work class would actually test those fixes. May 12 21:32:35 yes May 12 21:33:04 looking into that... May 12 21:33:11 with the 'putting staging under package management' project those problems should be gone May 12 21:33:26 * koen pimps his google SoC project May 12 21:33:43 and thanks all for explaining this, another part of voodoo unravelled May 12 21:33:46 <- erg bescheiden en subtiel May 12 21:34:07 koen: doe eens vertellen :-) May 12 21:34:14 hmmm, I need to go to sleep May 12 21:34:23 too much dutchisms May 12 21:34:25 * likewise i too curious May 12 21:34:30 s/i/is May 12 21:34:33 koen: what is your SoC project? May 12 21:34:42 ggilbert__: 'putting staging under package management' May 12 21:34:52 oh May 12 21:34:54 sweet May 12 21:35:18 That's something we've wanted forever :p May 12 21:35:27 * likewise nods May 12 21:35:43 koen: Not sure if you've considered this, but you might want something like dpkg-divert May 12 21:36:12 or maybe not May 12 21:36:34 munging wasn't happening on installed files, just during the installation process May 12 21:36:41 ggilbert__: I'll first make it work with ipkg, then make it general May 12 21:37:10 My mind is just thinking of the stuff like pkgconfig scripts that we munge May 12 21:37:25 but that would be covered already May 12 21:38:19 ggilbert__: what have you been munging during install? May 12 21:38:34 likewise: There are paths that needed changing May 12 21:39:19 __runcmd sed -e 's/^installed=yes$/installed=no/' -e '/^dependency_libs=/s,${WORKDIR}[[:alnum:]/\._+-]*/\([[:alnum:]\._+-]*\),${STAGING_LIBDIR}/\1,g' $dotlai >$destpath/$libname.la May 12 21:39:36 that's what base.bbclass does to the .la files May 12 21:40:26 My mind was going off in two different directions :p May 12 21:41:23 I was thinking there was a staging target that went and modified all of the instances in staging, which would suck come uninstall time May 12 21:41:28 but that's not the case May 12 21:41:31 koen: eek May 12 21:42:09 * likewise is not worthy May 12 21:42:39 likewise: sometimes you have to go to a bit of trouble to keep things in the host install from leaking into staging May 12 21:43:11 * JustinP reboots into a freshly updated oz354x e-image May 12 21:43:51 hrw|gone: are we going to get the boot logo with the new RC? May 12 21:43:59 * JustinP is waiting eagerly for it.... May 12 21:44:02 I like that new design May 12 21:45:01 JustinP: awaiting reply from the author, hrw is May 12 21:45:54 awaiting the "ok" to use his work for free? May 12 21:46:51 * JustinP watches his script spit out eap files May 12 21:46:59 I really ought to qemu this thing May 12 21:47:35 qemu runs about the same speed as a real device according to RP May 12 21:47:55 so it might not be a big win speedwise, but a nice polishy touch May 12 21:50:10 JustinP: http://ewi546.ewi.utwente.nl/tmp/add-new-splash.diff May 12 21:50:53 koen: cool. I was thinking more about the kernel splash, though ;-) May 12 21:51:02 http://ewi546.ewi.utwente.nl/tmp/hrw/t/logo.jpg May 12 21:52:26 probably run slower on my machine, actually....but would still be much better than forcing all of the users to run it on first boot....it takes 10 minutes or so May 12 21:52:29 on my spitz May 12 21:52:59 ~cheer for non-crashing xtscal May 12 21:53:01 * ibot cheers for non-crashing xtscal on May 12 21:54:05 * JustinP prepares to push newest e to oz354x May 12 22:10:51 03justinp 07org.oe.oz354x * r1ed43f7b... 10/ (62 files in 6 dirs): efl, e17: sync with .dev (also adds a patch which was previously missing) May 12 22:18:00 zaurusd isn't shutting down my spitz.... May 12 22:26:49 hrw|gone: changes pushed. I'd appreciate e images for RC3. :-) May 12 22:27:37 mickey|tv: could you commit the blackbox package? May 12 22:33:16 How come installing the icewm package doesnt include downloading the xfbdev package? May 12 22:33:46 RP: why is suspend-on-hinge-close turned off for zaurusd in .oz354x? May 12 22:34:08 RP: I foudn the commented lines and uncommented them, works just like it seems it should... May 12 22:50:06 g'nite all May 12 23:01:32 I want my own personal commit-authorized-oe-hacker , that way I can just push a button and watch him go and commit stuff. May 12 23:13:01 And I would call him Timmy May 12 23:13:07 ~lart Timmy May 12 23:13:08 * ibot lowers Timmy's priority May 12 23:35:38 ~lart timmy May 12 23:35:38 * ibot installs PocketPC on timmy's PDA May 12 23:36:35 KenJlime: you can be an OE pusher, just put enought patches in the bugtracker and ask for commit access May 12 23:47:55 JustinP: Yeah, will bother koen/mickey about it once they wake up.My patches are all mostly superh relative, so shouldnt risk breaking that much. May 13 00:49:26 re May 13 01:51:40 exit May 13 02:05:38 zwelch: ping May 13 04:28:10 snow4all May 13 04:28:39 snowball: ping May 13 04:29:15 oops - wrong channel May 13 04:47:03 03mwester 07org.oe.dev * rc4575033... 10/packages/linux/ (2 files in 2 dirs): Unslung: unslung-kernel pl2303.c add patch for modem control signals May 13 04:47:08 03mwester 07org.oe.dev * re604e0b9... 10/packages/nslu2-binary-only/unslung-rootfs/NOTES: Unslung rootfs update changelog in NOTES May 13 05:37:49 anyone know of a good web based cd burning tool? May 13 05:39:44 web based cd burning tool ... May 13 05:40:00 how would it know where your hardware is? May 13 05:40:26 and would you really trust an app with that much access to your system? May 13 05:41:31 emte: I trust Joerg Schilling implicitly. May 13 05:41:44 haha May 13 05:42:02 njs, i mean a web based app with access to your system hardware layer May 13 05:42:19 Jörg "You didn't mean to type that thing you did -- really you didn't" Schilling May 13 05:42:53 mithro: do you mean one you can download free on the net? web browsers don't get hardware access.... May 13 05:43:14 JustinP, there are ways May 13 05:43:23 but in the end would you trust it May 13 05:45:29 well....we download such programs and use them... May 13 05:45:39 but that's choosing to trust something May 13 05:45:52 its a bit different May 13 05:45:52 no, I don't want my web browser having hardware access May 13 05:46:15 your granting access to your ram and drives May 13 05:46:31 ( since you still need to cache somewhere) May 13 05:46:48 not need, just want ;) May 13 05:47:05 no, you need to May 13 05:47:26 data transfer of what you want to burn etc May 13 05:47:50 its an interesting topic anyway May 13 05:48:08 early versions of flash allowed full system access May 13 05:48:20 and we all know java still can May 13 05:49:15 but there are some protection systems from evil users now May 13 05:49:46 unless you are doomed to use windows ... since you will never get protection from MS themselves :P May 13 06:37:35 morning May 13 07:42:30 it May 13 09:35:21 hello, I build an oz354x bootstrap-image for my poodle and detected, that the /home partition at /dev/mtdblock3 isnt mounted. is it by intention? **** ENDING LOGGING AT Sat May 13 09:59:56 2006