**** BEGIN LOGGING AT Mon Jul 18 23:59:57 2005 Jul 19 00:03:28 is there a way to exclude minimo from the build ? Jul 19 00:36:29 morning all Jul 19 00:38:08 morning Jul 19 00:43:17 is vanille.de down again ? Jul 19 00:44:16 ade|desk: looks like Jul 19 00:46:12 hi hrw|work Jul 19 00:48:50 morning Jul 19 00:50:01 morning Jul 19 01:04:48 moring proti bluelightning XorA alan_ hrw|work etc Jul 19 01:06:23 hello ade|desk Jul 19 01:13:58 morning ade|desk Jul 19 01:31:00 hy ade|desk, proti, alan_ Jul 19 01:46:43 NOTE: package pwmpi-2.1.16-r0: task do_build: completed Jul 19 01:48:07 morning Jul 19 01:48:53 hi mickeyl Jul 19 01:50:03 heh.. vanille.de failed again Jul 19 01:51:18 mickeyl: morning. Jul 19 01:51:39 Harg! vanille.de closed while pulling. Jul 19 01:51:41 you and your killer clients :/ Jul 19 01:51:51 * mickeyl restarts server Jul 19 01:51:59 i think something's flaky in the netsync implementation Jul 19 01:52:05 mickeyl: Can't we just rsync a copy of the db ? Jul 19 01:52:06 probably need to submit a bug report Jul 19 01:52:20 proti: nah Jul 19 01:52:30 too bad Jul 19 01:52:30 proti: we have a problem with bitbake (again) Jul 19 01:52:37 proti: it seems that the collections fix broke something Jul 19 01:52:41 how to reproduce: Jul 19 01:52:48 set MACHINE = "spitz" Jul 19 01:53:26 then see in bitbake -i that the per package information contains PREFERRED_PROVIDERS "virtual/kernel:openzaurus-pxa27x" Jul 19 01:53:31 while the outer environment doesn't contain it Jul 19 01:53:36 hence we can't build it :/ Jul 19 01:53:46 if you have a moment, It would be great if you could look into it Jul 19 01:54:01 mickeyl: Ok. Did you have a testcase handy ? Jul 19 01:54:07 sure Jul 19 01:54:10 MACHINE = "spitz" Jul 19 01:54:14 bitbake -i Jul 19 01:54:19 getvar PREFERRED_PROVIDERS Jul 19 01:54:25 peek nano PREFERRED_PROVIDERS Jul 19 01:54:26 monotone: successful exchange with ewi546.ewi.utwente.nl Jul 19 01:54:27 uf Jul 19 01:54:27 that's it Jul 19 01:54:48 mickeyl: What do you get now, and what is the expected value ? Jul 19 01:57:40 proti: major problem is that the output of getvar and peek nano varies Jul 19 01:57:54 peek nano contains what we expect ("... virtual/kernel: openzaurus-pxa27x") Jul 19 01:57:56 getvar doesn't Jul 19 01:58:12 OK. it's enough for me. I'll have a look. Jul 19 01:58:16 thank you Jul 19 01:58:53 03mickeyl 07org.oe.dev * rcb61d6d8... 10/packages/zaurus-updater/tosa/updater.sh: zaurus-updater: recognize zimage.bin for tosa Jul 19 01:58:57 03jbowler 07org.oe.nslu2-linux * r0d839000... 10/packages/pcre/pcre_4.4.bb: Jul 19 01:58:57 Remove timestamp issues caused by the compilation of dftables external Jul 19 01:58:57 to the makefile (doesn't work when using ccache on a fast system because Jul 19 01:58:57 it attempts to rebuild dftables within the makefile). The change fixes Jul 19 01:58:57 the original problem that dftables was compiled by default with the target Jul 19 01:58:58 compiler. Jul 19 01:59:01 03hrw 07org.oe.dev * r86dd0e90... 10/packages/kdepimpi/ (4 files): updated kdepimpi to 2.1.16 Jul 19 01:59:05 03jbowler 07org.oe.nslu2-linux * r94d9b95f... 10/packages/bind/bind-9.3.1/lib_dns_Makefile.in.patch: Jul 19 01:59:05 Makefile.in was patched to change ${CC} to cc for something which needed Jul 19 01:59:06 BUILD_CC, but the target CFLAGS were still there, preventing the target Jul 19 01:59:08 cflags containing anything which only works with TARGET_CC. Fixed to use Jul 19 01:59:10 BUILD_CC, BUILD_CFLAGS and BULD_LDFLAGS - still a hack because these come Jul 19 01:59:12 in 'by magic' from the environment, but that's fine in OE builds because Jul 19 01:59:14 they are always set by the do_compile script. Jul 19 01:59:38 I like logs of nslu2 people - lots of info Jul 19 01:59:43 *nod* Jul 19 01:59:46 that's how it should go Jul 19 02:02:29 mickeyl: What distro do you use with the testcase ? Jul 19 02:02:38 proti: openzaurus-3.5.4 Jul 19 02:03:20 I am puzzled do_fetch on any minimo build for me ends up calling "pkg-config libIDL-2.0 --cflags" and locking up Jul 19 02:44:51 morning Jul 19 02:54:13 hi reenoo Jul 19 02:56:59 hey hrw|work Jul 19 03:05:59 wow, http://pastebin.com/316105. Jul 19 03:06:09 I hate monotone. Jul 19 03:06:19 mickeyl: ping Jul 19 03:07:10 looks good Jul 19 03:07:12 where's your problem? Jul 19 03:07:26 mickeyl: Elapsed time : 30min Jul 19 03:07:44 so what? 11 new revisions took some time Jul 19 03:07:52 mickeyl: Found your problem btw. Jul 19 03:08:00 oh that's good news Jul 19 03:08:04 was it due to the collection fix? Jul 19 03:08:15 mickeyl: There is no problem. Jul 19 03:08:26 hmm Jul 19 03:08:38 mickeyl: When going to interactive, you have to call update_data before. Jul 19 03:09:03 Else, the getvar returns non updated value. Jul 19 03:09:22 hmm i don't think the problem was limited to interactive mode Jul 19 03:09:36 but let me try that again Jul 19 03:09:52 mickeyl: http://pastebin.com/316131 Jul 19 03:10:06 I added line 8. Jul 19 03:11:23 And getvar shows me : http://pastebin.com/316132 Jul 19 03:11:36 hmm, k Jul 19 03:11:38 thanks Jul 19 03:11:51 mickeyl: The problem was already there. Jul 19 03:12:31 bbl, lunch Jul 19 03:26:13 03mickeyl * r288 10bitbake/bin/bitbake: update configuration data before entering the shell. patch courtesy Proti - thanks. Jul 19 03:32:19 03mickeyl 07org.oe.dev * r1e6a8e9d... 10/packages/parted/parted_1.6.23.bb: add GNU parted 1.6.23 - patch courtesy Pigi Jul 19 03:32:25 03mickeyl 07org.oe.dev * r189d3787... 10/packages/flite/ (flite-1.2/configure-with-audio.patch flite_1.2.bb): Jul 19 03:32:26 flite: Jul 19 03:32:26 - add configure-with-audio patch courtesy Patrick Ohly Jul 19 03:32:26 - reorder metadata Jul 19 03:32:26 - set Patrick Ohly as maintainer of this package Jul 19 03:43:31 03law 07org.oe.dev * r5b7b4ed7... 10/packages/gpsd/ (gpsd_2.26.bb gpsd_2.28.bb): updated gpsd to version 2.28 Jul 19 03:43:36 03mickeyl 07org.oe.dev * rd9ca92ef... 10/packages/lzma/lzma-native_4.17.bb: Jul 19 03:43:36 put some sense into lzma-native Jul 19 03:43:36 NOTE: NEVER EVER call 'make' directly in OE - use oe_runmake and adjust CFLAGS as necessary Jul 19 03:48:40 03jbowler 07org.oe.nslu2-linux * rca6fa63e... 10/conf/machine/nslu2.conf: Modified FULL_OPTIMIZATION for NSLU2 Jul 19 03:53:45 03jbowler 07org.oe.nslu2-linux * r3ee680bb... 10/packages/openslug-init/ (4 files in 2 dirs): Jul 19 03:53:45 slugbug 206: fix the tmpfs directories in NFS Jul 19 03:53:45 slugbug 207: fix the flashdisk mount so that it works in the presence of Jul 19 03:53:45 an existing readonly mount (turnup, like reflash, now removes all existing Jul 19 03:53:45 flashdisk mounts) Jul 19 04:03:31 Hi everyone Jul 19 04:04:55 mickey|lunch: It seems keys get automagically added to dbs Jul 19 04:06:25 OT: Trying to start the whole OE setup via a shell script (setting environment vars, etc.). I have export VAR=VALUE and I try to cd into the right directory so I can compile right away. But neither the export nor the cd seem to have any effect after the script ends. What am I doing wrong? Jul 19 04:06:53 Laibsch: source blah.sh Jul 19 04:06:59 you have to source the script Jul 19 04:07:05 . blah.sh Jul 19 04:07:32 ah, XorA was ahead of me :) Jul 19 04:07:35 Thanks guys. Jul 19 04:08:19 * hrw|work has setup-oe.sh for it Jul 19 04:09:05 03law 07org.oe.dev * r55dc1bb3... 10/packages/gaim/ (files/desktop-name_1.4.0.patch gaim_1.4.0.bb): added gaim 1.4.0 Jul 19 04:09:50 shit.. my home machine hang Jul 19 04:10:22 hrw|work: I always though hooking mains ontrol to a phoneline would be awesome Jul 19 04:11:09 XorA: I need ethernet watchdog Jul 19 04:11:40 aaargh Jul 19 04:11:45 ~lart rembours Jul 19 04:11:45 * ibot breaks out the Hoover and sucks up rembours Jul 19 04:12:10 hrw|work: that would work as well Jul 19 04:13:01 * koen grumbles a bit, and get's some money to pay the mailman when he comes back tomorrow Jul 19 04:16:27 http://www.fooishbar.org/blog/tech/libtool-2005-07-19.html Jul 19 04:16:36 * koen agrees with daniels Jul 19 04:20:30 getvar PREFERRED_PROVIDERS should print virtual/kernel, right? Jul 19 04:22:25 proti, mickey|lunch: any idea on this: http://pastebin.com/316154 ? Jul 19 04:22:39 koen: r288 fixed that iirc Jul 19 04:22:48 it didn;t Jul 19 04:23:17 bitbake:/data/build/oe/bbsvn/trunk# svn up Jul 19 04:23:17 At revision 288. **** BEGIN LOGGING AT Tue Jul 19 05:08:40 2005 Jul 19 05:10:40 http://ewi546.ewi.utwente.nl/maps/OE.php now has updated project icons Jul 19 05:14:42 <__law__> koen, looks nice :-) Jul 19 05:24:58 * chouimat is away: Desktop Developers' Conference aka gnome stuff :( Jul 19 06:11:15 hi reenoo_ Jul 19 06:11:25 hey pb_ Jul 19 06:12:39 I have set SOURCEFORGE_MIRROR in conf/local.conf. Compilation again failed because that particular mirror does not have the lex sources. Did I do something wrong or is this a bug where bitbake fails to just look for a different mirror? Jul 19 06:13:08 My particular mirror is mesh and http://mesh.dl.sourceforge.net/sourceforge/lex/flex-2.5.31.tar.bz2 does not exist. Jul 19 06:13:32 try heanet Jul 19 06:14:34 koen: Of course I know that I can get the file somewhere else. I was talking more about the general problem this seems to suggest with specifying a mirror in local.conf. Jul 19 06:14:49 Laibsch: default is heanet as it contain all sources Jul 19 06:15:07 BTW, can I just resume the compilation or do I have to start afresh? Jul 19 06:15:24 Laibsch: the problem is at sourceforge Jul 19 06:15:35 we can't help if the mirror don't contain everything Jul 19 06:15:39 hrw|work: OK, nice to know. I'll change to it for now. But that kind of negates the meaning of a mirror, doesn't it? Jul 19 06:16:04 blame sourceforge Jul 19 06:16:37 koen: Well, I think it makes sense for them to not require every mirror to really carry everything. Jul 19 06:17:07 they list them as 'mirrors', so they should contain everything Jul 19 06:17:48 Well, if you download from the web page, they will list only the mirrors that contain the actual sources, so I think their approach is valid and I would *not* blame them. Jul 19 06:18:03 Anyway, can I resume the compilation or not? Jul 19 06:18:11 not meaning to start anew. Jul 19 06:18:17 'not' meaning to start anew. Jul 19 06:18:44 try and see Jul 19 06:19:06 as I can't understand how you can 'resume' compiling something you said you don't have the source of Jul 19 06:19:58 I meant the stuff that was compiled before the fetch for lex failed. Jul 19 06:21:32 Will that be fetched and compiled again? Wasting time and bandwidth in the process. Jul 19 06:21:55 try and see Jul 19 06:26:38 *sigh* Jul 19 06:26:44 "openzaurus gui" Jul 19 06:26:46 http://www.oesf.org/forums/index.php?showtopic=13938 Jul 19 06:26:53 stinking zaurus using retards Jul 19 06:28:27 koen: don't read forums. problem solved. Jul 19 06:28:57 yeah Jul 19 06:29:13 koen: Well, it seemed to start anew last time so I thought maybe I made a mistake, forgot to set a switch, $whatever. So I thought, I'd ask. Jul 19 06:29:34 koen: you sound like you need beer Jul 19 06:29:55 koen: What is the supposed result of 'getvar PREFERRED_PROVIDERS' Jul 19 06:30:02 from your pastebin ? Jul 19 06:30:17 proti: at least something with virtual/kernel Jul 19 06:32:04 proti: as you can see with bbread h3900.conf does PREFERRED_PROVIDER_virtual/kernel_kernel24 = handhelds-pxa Jul 19 06:32:14 and it worked before Jul 19 06:43:40 koen: Where the PREFERED_PROVIDER_virtual/kernel_kernel24 is added to PREFERRED_PROVIDERS ? Jul 19 06:44:45 proti: it isn't. Jul 19 06:45:23 pb_: So, how can it be present in PREFERRED_PROVIDERS ? Jul 19 06:45:27 It can't. Jul 19 06:45:44 pb_: That's what I suspected. Jul 19 06:45:46 well, not unless you actually set PREFERRED_PROVIDERS to that value. Jul 19 06:46:10 Nothing in bitbake modifies PREFERRED_PROVIDERS implicitly. Jul 19 06:46:49 koen: Could you detail why this value should be present in PREFERRED_PROVIDERS ? Jul 19 06:47:07 I assumed every PREFERRED_PROVIDER was added to that Jul 19 06:47:12 but anyway: Jul 19 06:47:13 BB>> getvar PREFERRED_PROVIDER_virtual/kernel Jul 19 06:47:13 None Jul 19 06:47:21 Of course Jul 19 06:47:37 This is PREFERRED_PROVIDER_virtual/kernel_*kernel24* Jul 19 06:47:53 or kernel26 Jul 19 06:48:44 proti: kernel24 is in OVERRIDES Jul 19 06:50:06 koen: line 547 of bitbake should read : bb.data.update_data(self.configuration.data) Jul 19 06:50:29 This assure that the override is taken care of before going to interactive. Jul 19 06:51:41 it doesn't work in non-interactive mode too Jul 19 06:52:23 proti: and line 547 is exactly what you mentioned Jul 19 06:52:25 What is your ovverride and can you do a testcase for it (gather conf files needed to show bad behaviour). Jul 19 06:52:54 set MACHINE=h3900 and bitbake virtual/kernel Jul 19 06:53:17 both bitbake and OE are the latest version and unmodified Jul 19 06:53:25 I will try now. Jul 19 06:57:15 koen: ok reproduced. I'll tell you when I found smth. Jul 19 06:57:26 thanks Jul 19 07:08:04 03law 07org.oe.dev * rb9180e61... 10/packages/sylpheed/ (2 files in 2 dirs): added sylpheed 2.0.0beta6 Jul 19 07:08:08 03law 07org.oe.dev * ra84100b1... 10/packages/gpgme/gpgme_1.0.3.bb: added ggme 1.0.3 (required for sylpheed 2.0.0beta6) Jul 19 07:11:46 * mickeyl freshly baken TomTomGo 500 owner Jul 19 07:12:13 hi Jul 19 07:12:17 cheers CoreDump|home Jul 19 07:14:47 hi CoreDump Jul 19 07:15:08 hey hrw|work mickeyl :) Jul 19 07:16:58 mickeyl: when do you add OE support for it ;) Jul 19 07:20:09 koen: probably never. I want _one_ device where I'm just a stinkin' user :D Jul 19 07:22:41 hi Jul 19 07:23:16 hey Bernardo Jul 19 07:23:22 hi Bernardo Jul 19 07:24:19 hi Bernardo Jul 19 07:24:39 what is tomtomgo500? gps device? Jul 19 07:25:35 ya. it's a standalone navigation device. (based on linux, btw.) Jul 19 07:27:11 ~curse gpsd Jul 19 07:27:11 May the fleas of a thousand camels infest your most sensitive regions, gpsd ! Jul 19 07:28:04 You better also lart the lazy folks who insist on ignoring that a software is maintained and developed. Jul 19 07:30:28 ahem Jul 19 07:30:31 do_package() { Jul 19 07:30:31 base_do_package Jul 19 07:30:31 if (bb.data.getVar('INHIBIT_PACKAGE_STRIP', d, 1) != '1'): Jul 19 07:30:31 bb.build.exec_func('do_strip_modules', d) Jul 19 07:30:31 } Jul 19 07:30:57 no wonder that b0rks in a bash script :) Jul 19 07:31:24 nod. i think there's a missing 'python' in there Jul 19 07:31:34 i wonder why it works despite that Jul 19 07:31:45 because actually, it does. we are building distros with OE :) Jul 19 07:31:58 it doesn't for me Jul 19 07:32:25 sigh Jul 19 07:32:33 looks like i have to trash tmp Jul 19 07:38:44 mickeyl: that only happens if you forget to set a DISTRO. Jul 19 07:38:56 which, almost by definition, you don't forget to do when you are building a distro Jul 19 07:39:28 pb_: ah right. Jul 19 07:39:56 or, more specifically, it happens if you don't inherit package_ipk or the like. Jul 19 07:40:02 "Linus on changing the default I/O scheduler to CFQ: It better not eat anybody's disks, because when someone loses his porn collection, I hear about it for years." Jul 19 07:40:07 that's definitely a bug, but nobody has cared enough to fix it yet Jul 19 07:40:14 pb_: which jffs2-summary patch are you talking about on alkml? Is it something we should apply to our Zaurus kernels to improve performance? Jul 19 07:40:25 koen: hehe Jul 19 07:40:58 mickeyl: http://www.inf.u-szeged.hu/jffs2/mount.php Jul 19 07:41:02 that one, I think Jul 19 07:41:19 iirc that one is already in Jul 19 07:41:27 otherwise mounting 128MB is *very* slow Jul 19 07:41:29 oh - interesting. never seen that Jul 19 07:41:55 i wonder if that applies to our embedix kernels Jul 19 07:43:29 hmm probably not since it's based on a more or less recent cvs Jul 19 07:43:40 koen: heh Jul 19 07:46:11 mickeyl: you could look into updating the mtd code for 2.4.18 and backporting the summary patch Jul 19 07:46:21 but that would probaly more trouble than it's worth Jul 19 07:46:29 *nod* Jul 19 07:46:43 besides I've promised to never look at 2.4.-embedix again and I like to keep my promise Jul 19 07:46:53 We don't use the summary patch even in 2.6 yet Jul 19 07:47:09 mickeyl: very wise Jul 19 07:47:27 RP: ah, and we don't have and partition >70MB iirc Jul 19 07:47:43 /and/any/ Jul 19 07:48:23 koen: correct Jul 19 07:52:47 ~whois koen Jul 19 08:02:05 has anybody else got a GCC internal error while trying to build x with OE? Jul 19 08:02:38 amd64? Jul 19 08:03:05 I dont even count ICE from gcc - usually they are non repeatable on my system so next attempt to build goes ok Jul 19 08:03:10 Nah, i386. Athlon to be exact Jul 19 08:03:27 I tried this one a couple of times and got the same error Jul 19 08:03:48 could you pastebin the error? Jul 19 08:03:58 and wipe tmp, update OE, etc Jul 19 08:04:29 I tried wiping tmp, didn't change anything, although I didn't wipe my sources directory Jul 19 08:06:50 koen: http://pastebin.ca/18060 Jul 19 08:08:40 03jbowler 07org.oe.nslu2-linux * r37a0ff9e... 10/ (conf/distro/openslug.conf packages/linux/nslu2-kernel.inc): Jul 19 08:08:40 Support for disabling irq debug messages - slugbug 109. Jul 19 08:08:40 This allows a simple local.conf change to re-enable the debug messages Jul 19 08:08:40 simply by setting CMDLINE_DEBUG to "" Jul 19 08:08:44 koen: I see what is wrong. Jul 19 08:09:25 hrw|work: if you get non-repeatable ICEs, your computer is broken Jul 19 08:11:06 pb_: I know that - no time to check what is wrong this time Jul 19 08:13:18 koen: Want to know what is wrong ? Jul 19 08:15:20 don't tease us. speak up! :) Jul 19 08:15:26 A nice little riddle. If I do 'OVERRIDES_append="foo:" and call update_data. foo is included in the possible OVERRIDES values. Must call update_data twice. Jul 19 08:15:40 d'oh Jul 19 08:15:43 foo is NOT included in the OVERRIDES. Jul 19 08:15:47 I mean. Jul 19 08:16:39 proti: I see Jul 19 08:17:20 so, what can we do to fix that without introducing more regressions ? Jul 19 08:17:23 koen: Fix is call update_data twice or fix the semantics of OVERRIDES. Jul 19 08:17:28 BigAl: any reason why you want to build xorg for arm? Jul 19 08:17:32 mickeyl: Double the call to update_data. Jul 19 08:17:44 ugly. Jul 19 08:17:46 hmm Jul 19 08:17:54 mickeyl: update_data probably ought to iterate until the dataset stabilises. Jul 19 08:18:44 koen: I did a 'bitbake gpe-image' and it built that automatically Jul 19 08:19:11 pb_: And then, I'm sur I'm able to build a dataset that leads to endless loop. Jul 19 08:19:18 hmm Jul 19 08:19:24 * mickeyl thinks about that while wandering home Jul 19 08:19:29 BigAl: you MACHINE should set the xserver to kdrive Jul 19 08:19:31 03nail 07org.oe.nslu2-linux * rfd2bac75... 10/conf/distro/openslug.conf: Split the feeds into cross/native and ${DISTRO_VERSION}/unstable Jul 19 08:19:33 Maybe not. Jul 19 08:19:51 mickeyl: Fix is don't allow override of override. Jul 19 08:20:09 proti: I don't think it is possible to engender an infinite loop. Jul 19 08:20:44 afaik, the overrides must eventually converge on a stable configuration. Jul 19 08:21:25 pb_: No, you're right. But update_data is already the #1 of cpu consumption. Iterating until it stabilize will kill perfs a little bit more. Jul 19 08:22:08 Mm, that's true. Jul 19 08:24:30 I think that OVERRIDES_prepend="${KERNEL}:" is in fact : OVERRIDES += "${KERNEL}:". or OVERRIDES = "${KERNEL}:${OVERRIDES}" Jul 19 08:24:53 koen: Weird.. I wonder why it's not. Might be time to update bitbake again. Jul 19 08:25:43 proti: the latter would be an error; you can't define a variable in terms of itself. You'd have to use := in that case. Jul 19 08:26:05 the former will insert a stray space, which might or might not be a problem Jul 19 08:27:27 pb_: Yes, := indeed. But it is not my point. OVERRIDES_prepend or OVERRIDES_append should not be allowed or treated in a special case. Jul 19 08:28:16 proti: I think one could argue that being able to add things to OVERRIDES conditionally would be useful. Jul 19 08:28:34 and the only way to do that, of course, is with OVERRIDES_prepend_ Jul 19 08:29:06 I guess it wouldn't be the end of the world if that was outlawed. Jul 19 08:30:23 Hum, this not the way bitabke do the things. bitbake doesn't allow conditionnal definitions. only overides. And IMHO, it is good. Jul 19 08:30:47 right, that was my point. Jul 19 08:31:12 if you want to have something added to OVERRIDES conditionally, the only way to accomplish it is to manipulate OVERRIDES using another override. Jul 19 08:31:49 so, if you disallow overrides of OVERRIDES, it would actually lead to a (small) loss of functionality. Jul 19 08:34:13 Ok, so, what to do ? Jul 19 08:36:09 dunno. check with mickey and zecke, see what they think. Jul 19 08:36:20 I think the loss of functionality is probably tolerable. Jul 19 08:36:57 mickeyl: ping Jul 19 08:37:18 How, I forgot he left to home. Jul 19 08:45:13 03rw 07org.oe.dev * r974e37be... 10/packages/classpath/ (classpath_0.16.bb classpath_0.17.bb): classpath: version bump to bugfix release 0.17 Jul 19 09:03:26 cu Jul 19 09:41:26 hi france_ Jul 19 09:42:05 pb_: hi Jul 19 09:55:40 yay, a bug in monotone...I can't merge.... Jul 19 09:55:48 I suppose I'll have to manually merge.... Jul 19 09:56:41 re Jul 19 10:22:10 * Luke-Jr wonders why OE gets people who can't read Jul 19 10:23:15 03nail 07org.oe.nslu2-linux * r0788484c... 10/packages/samba/samba_3.0.14a.bb: Added smb.conf to CONFFILES Jul 19 10:26:57 Luke-Jr: by installing linux you become an expert on *everything*, so why read? Jul 19 10:30:41 Like-Jr: to what are you referring? Jul 19 10:47:09 JustinP: reading 'glib' as 'libc', 'glibc' as 'gcc', etc Jul 19 10:47:21 JustinP: OE mailing list Jul 19 10:54:11 oh that...yeah, that was annoying Jul 19 11:04:40 mickeyl: is mt still slow? Jul 19 11:06:36 if you sync twice a day it's faster as bk with it's consistency check Jul 19 11:07:23 I'm just kidding Jul 19 11:08:01 if you want to sync after week without syncing then is is fscking slow Jul 19 11:23:08 bleh Jul 19 11:23:13 palm's developer site sucks ass Jul 19 11:23:29 03jbowler 07org.oe.nslu2-linux * r3a43238b... 10/packages/openslug-init/ (openslug-init-0.10/functions openslug-init_0.10.bb): Jul 19 11:23:29 unmountflash would make /dev/mtdblock4 on / ro, this only affects a turnup Jul 19 11:23:29 when the flash file system is mounted on /, but it causes the edit of the Jul 19 11:23:29 /linuxrc to fail. Jul 19 11:24:03 proti: your fix only fixes shell mode Jul 19 11:27:15 koen: please ignore my last mail Jul 19 11:27:25 koen: wrong email address. right key will follow Jul 19 11:27:43 the key from _law_? Jul 19 11:27:47 no Jul 19 11:27:50 _law_ is ok Jul 19 11:27:52 coredump was wrong Jul 19 11:27:55 (sent 1 minute ago) Jul 19 11:28:03 hh.org mailserver seems to lag anyway Jul 19 11:28:09 hi Jul 19 11:28:09 so it probably arrives the same time as the fixed key Jul 19 11:28:39 yo tubbie|home Jul 19 11:29:37 mickeyl: it seems keys get autoimported once a cert exist signed by that key Jul 19 11:29:58 koen: oh is that so? interesting Jul 19 11:30:15 ok. time to go for bread Jul 19 11:30:32 mickeyl: it makes sense that it does so Jul 19 11:30:38 koen|ewi: righto. we don Jul 19 11:30:40 enough hacking for today Jul 19 11:30:48 t want revs from not existing people Jul 19 11:30:51 hrw: enjoy Jul 19 11:31:06 bye hrw Jul 19 11:31:10 hrw: smagsnego (or something like that) Jul 19 11:31:29 koen: thx Jul 19 11:31:40 and its 'smacznego' Jul 19 11:32:01 mickeyl: but I'll add the key anyway if vanille.de is down and coredumps wants to push his first cert :) Jul 19 11:32:46 * CoreDump|home is creating a new key Jul 19 11:33:53 koen: yep Jul 19 11:33:55 ibot: bon appetit is smacznego Jul 19 11:33:55 okay, mickeyl Jul 19 11:33:57 ibot: bon appetit is also Guten Appetit Jul 19 11:33:57 okay, mickeyl Jul 19 11:34:04 feel free to add more languages Jul 19 11:34:40 ibot: bon appetit is also Eet Smakelijk Jul 19 11:34:40 koen: okay Jul 19 11:35:54 * mickeyl goes and plays with his TomTomGo500 Jul 19 11:35:59 l8er Jul 19 11:36:50 ibot: bon appetit is also God Appetitt Jul 19 11:36:50 okay, NAiL Jul 19 11:40:37 koen: keys: yes, it seems that monotone views the key as authentication, not as a security measure. The security has to be implemented in the LUA hooks in the server monotonerc. Jul 19 11:45:40 jbowler-away: how so? can you add key checksums there? Jul 19 11:48:41 The LUA hooks take an 'identity' - for me that's jbowler@nslu2-linux.org - and permit/deny write access based on this. Jul 19 11:48:57 The identify is secure because it is just a string name for a public key stored in the server. Jul 19 11:49:34 My monotone encrypts using my private key, and the OE monotone server decrypts using it's copy of my public key. Jul 19 11:49:34 well Jul 19 11:49:56 koen just said the server auto-import keys Jul 19 11:50:05 imports even Jul 19 11:50:37 Correct. So the server auto-imports identities, but only from authorities which already have a known identity (therefore are trusted). Jul 19 11:51:22 huh? you can use any mail address you want when generating a key Jul 19 11:51:36 So? Jul 19 11:51:43 if it matches the server will accept it no matter who you actually are Jul 19 11:51:57 If it matches what... Jul 19 11:52:53 the identity string on the server Jul 19 11:53:12 I still don't understand - which identity string on the server? Jul 19 11:55:45 If I attempt to write to monotone.vanille.de I have to send the hash of my public key, and I have to have the private key to encrypt the data. monotone.vanille.de has to (1) have that public key (with that has) (2) permit the corresponding identity to write (3) be able to decrypt the stream with that public key. Jul 19 11:56:00 s/has)/hash)/ Jul 19 11:57:36 I can add any number of identities to montone.vanille.de (because I do have write access as 4ce8a13ed7cd0e321e101f1e4e4d5aad352b1a85), but those identities can be blocked at (2). Jul 19 11:58:28 I can't change an existing identity on monotone.vanille.de - it will whine like mad if I pass it a certificate with a changed key (and it will ignore it). Jul 19 11:59:08 hmm. ok. Jul 19 12:00:19 But step (2) is very very important. Because if monotone.vanille.de just accepts writes from any identity whose public key it has (returns true unconditionally in the LUA hook which implements 2) then sooner or later an untrustworthy individual will end up with write access... Jul 19 12:01:27 well. I think the concept of auto-importing keys is flawed in general Jul 19 12:01:44 no matter how you describe it Jul 19 12:02:32 Its the same as the 'web of trust' stuff. It doesn't prove someone is nice, just that they are foo@bar.org Jul 19 12:02:49 I.e. it is authentication, not security. Jul 19 12:02:55 no it's not the same Jul 19 12:03:12 you have to explicitely import keys with say gpg Jul 19 12:03:48 Ok, true - it's splitting the concepts even more. Jul 19 12:04:25 All you can say, a priori, is that hacker@nslu2-linux.org is the same identity every time. Jul 19 12:04:53 So then you accept write access by explicitly adding hacker@nslu2-linux.org to the LUA hook. Jul 19 12:05:24 And you accept changes made by hacker@nslu2-linux.org by writing a trust LUA hook and explicitly accepting the certificates hacker generates. Jul 19 12:05:25 reenoo_: heya - have you been keeping track of Sven's recent work in Classpath? Jul 19 12:05:47 Twiun: the Gtk image stuff? Jul 19 12:06:04 reenoo_: more recent Jul 19 12:06:24 reenoo_: http://www.qc.physto.se/~sven/qtpeers.png Jul 19 12:06:24 Twiun: I'm afraid no. been pretty much out of the loop Jul 19 12:06:40 reenoo_: he pretty much gave up on gtk Jul 19 12:06:54 ... and for those wondering - I'm keeping an eye on these for Java on OE Jul 19 12:06:58 * reenoo_ shrugs Jul 19 12:08:01 Twiun: well. I'm certainly not going to work on anything qt Jul 19 12:08:12 Twiun: nice Jul 19 12:08:25 reenoo_: phylosophical or just time? Jul 19 12:08:37 reenoo_: btw, he's the only one working on those :) Jul 19 12:08:39 reenoo_: but Qt provides everything you miss from monotone Jul 19 12:08:48 reenoo_: everyone else is GTK bound Jul 19 12:09:02 zecke: surely it doesn't provide a garbage collector? Jul 19 12:09:26 Twiun: no garbage, no need to collect ;) Jul 19 12:09:34 Twiun: Qt being owned by TT is a bad idea by definition Jul 19 12:09:52 reenoo_: you still don't like qt4 even GPL'd? Jul 19 12:09:54 reenoo_: but Qt is mature, almost bug free, well supported, mature Jul 19 12:10:03 * Twiun senses a theme here Jul 19 12:10:22 :} Jul 19 12:11:08 * CoreDump|home chuckles Jul 19 12:11:17 Twiun: I don't like the implications of the TT connection. no matter what the license is. Jul 19 12:11:36 reenoo_: 'k - that's good - more choice Jul 19 12:18:14 Twiun: yeah, sure. no problem with someone else adding and maintaining a classpath-qt.bb Jul 19 12:19:16 * Twiun winks Jul 19 12:19:35 What implications are there exactly? Jul 19 12:19:39 it's under the gpl Jul 19 12:19:46 what else do you want? Jul 19 12:20:03 norwegen=urlaub Jul 19 12:20:24 sorry wrong window :) Jul 19 12:20:42 treke: copylefted to the fsf :) Jul 19 12:20:55 urlaub sounds nice do13 :) Jul 19 12:21:06 urlaub=vacation Jul 19 12:21:10 * chouimat|ibook is so tired of this stupid TT is evil crap Jul 19 12:23:47 tubbie|home: two weeks too late. I past my vacations. Jul 19 12:26:27 chouimat|ibook: I didn't say TT was evil. it's just that.. I mean, remember what happened after Be Inc went bancrupt. much of the code involved was GPLed but without the corporate backing... Jul 19 12:27:12 RP: Did you tested 2.6.13-rc3-mm1? Jul 19 12:27:15 and what would happen to say gtk if redhat were to go under? Jul 19 12:27:17 * zecke wonders what Gtk would be without $$$ Jul 19 12:29:14 treke: last time I checked RH wasn't the only company behind gtk Jul 19 12:29:33 no, but they are a pretty big force behind it Jul 19 12:31:17 reenoo_: if something bad happens to TT the last released version of Qt will be automaticaly under a BSD license ... the KDE project have the signed papers for it Jul 19 12:31:31 if trolltech were to go out of business, qt development would probably slow down signficantly until new people got back up to speed Jul 19 12:31:43 But it would likely go on Jul 19 12:32:46 Some of the trolltech developers would probably continue working on it Jul 19 12:33:01 treke: and the kde developers too Jul 19 12:34:10 chouimat|ibook: ah, interesting. I wasn't aware of that. Jul 19 12:34:20 so bankrupting TT would do the world a favour? Jul 19 12:34:38 ... Jul 19 12:34:39 * koen whistles innocently Jul 19 12:34:48 fucking idiots Jul 19 12:34:58 so easy :P Jul 19 12:35:23 heh Jul 19 12:35:50 koen bad Jul 19 12:36:21 dual licensing bad Jul 19 12:36:27 single license good Jul 19 12:36:41 Jul 19 12:36:48 reenoo_: http://www.kde.org/whatiskde/kdefreeqtfoundation.php Jul 19 12:36:49 koen: the "give koen all the money" license? Jul 19 12:37:08 Twiun: no, gpl or lgpl Jul 19 12:37:21 not the gpl till you want to make money license Jul 19 12:38:46 chouimat|ibook: ah. good to know Jul 19 12:39:04 I don't dislike qt or trolltech Jul 19 12:39:07 koen dual licensing is how to convert the virus GPL into a a symbiot... actualy if I write a framework I will also dual license it. lgpl is a stupidity ... if you want to work like a crazy on your stuff and a multinational decides to use it and you doesn't get any money of their millions they are making out of your stuff it's bad imho ... but then again if you don't need to eat it's another story Jul 19 12:39:10 I jsut resent the hack called qt/e Jul 19 12:39:49 tubbie|sg-atlant: season premiere? Jul 19 12:39:53 koen look like qt/e 4.0 will be a lot better (at least from what I saw in the prerelease) and i think you mean qtopia Jul 19 12:40:02 no, qt/e Jul 19 12:40:03 yes koen Jul 19 12:40:07 :) Jul 19 12:40:13 tubbie|sg-atlant: enjoy, it's very nice Jul 19 12:40:23 and BSG season 2 started too Jul 19 12:40:25 :) Jul 19 12:40:44 thx koen Jul 19 12:41:01 chouimat|ibook: I get sick of people saying qt/e is "faster" and "compatible" Jul 19 12:41:26 koen: It is compatible to itself Jul 19 12:41:33 :) Jul 19 12:41:40 koen: and it is faster thant the nonexistant GTK/E version Jul 19 12:41:47 hey zecke Jul 19 12:41:49 you mean gtk/fb Jul 19 12:42:14 koen: no Gtk/Embedded, I wanted a fair test between Gtk/E and Qt/E Jul 19 12:42:19 heh Jul 19 12:42:46 sady Gtk folks failed to deliever their attendee Jul 19 12:42:54 so Qt/E won and is faster and compatible Jul 19 12:43:21 IMHO only DirectFB belongs on Consumer Electronic Jul 19 12:56:59 how do I strip the keys from a monotone DB (to create a snapshot)? Jul 19 12:57:20 JustinP: dropkey? Jul 19 12:57:26 do13: I've not done anything with the Zaurus kernel for a while due to holidays etc Jul 19 12:57:32 db dropkey KeyID Jul 19 13:04:33 thanks Jul 19 13:04:40 now I have to figure out how to list the keys... Jul 19 13:04:49 JustinP: list keys? Jul 19 13:05:23 ibot: botmail for treke I'm sorry this turned into yet another flamewar. my only point was that I have my own personal doubts and therefore decided not to use/support Qt myself. I'm absolutely fine with the rest of the world using Qt. and chouimat's hints have certainly made things look more positive Jul 19 13:05:26 oh yeah....I meant to check list but forgot Jul 19 13:05:29 thanks again Jul 19 13:07:25 RP: http://www.do13.de/openzaurus/temp/linux-openzaurus_2.6.13-rc3-mm1.bb Jul 19 13:10:41 do13: Ok, thanks. I'd prefer to wait until after I've built and tested it before committing if you don't mind though :) Jul 19 13:13:34 RP: Yes. Please test this on your machine. On the tosa i have one failure on startup: "BUG: soft lockup detected on CPU#0!" with an adress from my nand driver but after this message the system works. Jul 19 13:14:52 do13: I'd suspect the arm smp code :-/ Jul 19 13:18:25 n8 Jul 19 13:30:59 awwwww. I missed out... Jul 19 13:31:41 hey Jul 19 13:32:23 hi Jul 19 13:34:36 hi ljp Jul 19 13:35:40 RP: wouldn't you want and SMP pda? Jul 19 13:35:59 RP: I bet sharp would make one and let lineo mess up the kernel Jul 19 13:37:10 RP: or reprogram the wifi on the 770 to be a FPU ;) Jul 19 13:43:22 anyone know a good mapping site for vienna? Jul 19 13:44:24 (wild guess) map24.at ? Jul 19 13:46:22 basically I'm trying to find out where this place is Jul 19 13:46:41 Speisinger Strasse 13 Jul 19 13:46:47 1130 Wien Jul 19 13:48:46 http://tinyurl.com/bbnkr Jul 19 13:48:48 hey zecke, you alive? Jul 19 13:49:02 * zecke checks Jul 19 13:49:11 mithro: not sure if that'll work in your browser though Jul 19 13:49:17 test failed Jul 19 13:51:30 koen: smp would be nice. It remains to be seen how much of a mess Sharp will make of it :) Jul 19 13:57:40 mithro: maybe this works better: http://tinyurl.com/9gndv Jul 19 14:00:08 RP: as long as they do not use MACH and let two Linux kernel run side by side Sharp can mess with it Jul 19 14:01:23 zecke: Please don't even think that :) Jul 19 14:04:30 anyone have a maemo image for zaurus? Jul 19 14:04:46 bitbake maemo-image Jul 19 14:04:56 my OE is too messed up now to build one Jul 19 14:05:17 good night ! Jul 19 14:05:22 cya Jul 19 14:05:42 03nail 07org.oe.nslu2-linux * r76ec428e... 10/packages/samba/samba_3.0.14a.bb: Back out CONFFILES change that only worked in my imagination Jul 19 14:07:38 maemo? Jul 19 14:07:41 what's that? Jul 19 14:08:19 no time to build Jul 19 14:08:28 * JustinP starts a build Jul 19 14:08:37 JustinP: maemo.org Jul 19 14:08:42 yeah, I'm there Jul 19 14:09:16 JustinP: http://oss.kernelconcepts.de/maemo/n770-hildonui.png Jul 19 14:10:32 I see...sorta Jul 19 14:10:43 so is this supposed to work on Zaurii then? Jul 19 14:11:29 it kinda works Jul 19 14:11:44 it is made for 800x480 screens Jul 19 14:11:59 so you 'miss' 160x480 pixels Jul 19 14:12:27 JustinP: on a c700: http://dominion.kabel.utwente.nl/koen/images/maemo-on-Z.png Jul 19 14:13:01 hey zer Jul 19 14:13:04 bah Jul 19 14:13:07 mithro: hey Jul 19 14:13:18 hey zecke Jul 19 14:15:59 do you know harlekin's email address? Jul 19 14:16:05 the handhelds.org doesn't like me Jul 19 14:16:51 Maximilian Reiss Jul 19 14:20:08 arg that map24 requires java Jul 19 14:20:15 and i don't have java installed Jul 19 14:20:55 try swtiching to the non-java version Jul 19 14:21:10 could someone take a screenshot and email it to me? :P Jul 19 14:21:11 koen: ah, so it's not really useful on the Z at all Jul 19 14:21:15 the "statisch" link in the upper right corner Jul 19 14:21:44 JustinP: with the current situation, no Jul 19 14:21:59 hrm, Jul 19 14:22:02 The recipient does not Jul 19 14:22:02 accept mails from 'mithis.com' over foreign mailservers. 550 5.7.1 Jul 19 14:22:02 According to the domain's SPF record your host '192.58.209.91' is not a Jul 19 14:22:02 designated sender. (in reply to RCPT TO command) Jul 19 14:22:43 mithro: SPF *eek* Jul 19 14:22:57 HH.org will be down for maintenance 8am EST to noon tomorrow. Jul 19 14:24:58 reenoo_: you are austrian? Jul 19 14:25:23 koen: What is wrong ? Jul 19 14:25:45 mithro: german Jul 19 14:25:54 proti: lots of stuff Jul 19 14:26:04 reenoo_: where in germany? Jul 19 14:26:19 koen: Well be more specific. Jul 19 14:27:07 mailinglist archives and the adminsystem are the big ones Jul 19 14:27:27 ldap and broken python scripts come after that Jul 19 14:27:55 mithro: northern part of .de. near Bremen Jul 19 14:28:12 The incorrect values are because nothing calls update_data for the configuration.data. Jul 19 14:29:16 In parsing the bb files the update_data is called twice setting the correct value. Jul 19 14:29:29 koen: So what is wrong with my fix ? Jul 19 14:29:54 it only work in the shell, not with 'bitbake virtual/kernel' Jul 19 14:30:10 reenoo_: tried click on the statisch link but it's still got the plugin picture Jul 19 14:30:12 so I can build stuff, but all my scripts are useless now Jul 19 14:31:01 koen: No, it is a different issue. If you want a ugly hack, duplicate the update_data line. Jul 19 14:31:13 ahh ah! Jul 19 14:31:21 * mithro gets it to change Jul 19 14:31:33 I did, and it only fixes the interactive mode, since it is in the interactive block Jul 19 14:32:12 Ah ok. Did this worked before and when ? Jul 19 14:32:16 now if only the picture would load :/ Jul 19 14:34:19 proti: it worked about a month ago Jul 19 14:35:18 reenoo_: do you think you could just email me a picture? :P i'm having a hard time trying to get one Jul 19 14:35:35 mithro: working on it... Jul 19 14:39:50 mithro: http://handhelds.org/~rw/tmp/ Jul 19 14:40:46 mithro: the [1] apparently marks the destination Jul 19 14:43:25 koen: Move the 2 update_data after the line 599 : self.parseConfigurationFile( os.path.join("conf", "bitbake.conf") ) Jul 19 14:44:20 Laibsch: still there? Jul 19 14:44:21 * koen does so Jul 19 14:45:12 proti: that fixes it, thanks Jul 19 14:45:40 koen: You're welcome but it is so ugly, don't tell anyone I'm the author of this. Jul 19 14:46:11 * reenoo_ rotfl Jul 19 14:46:21 question about directory structure... Jul 19 14:46:31 * koen runs another image build Jul 19 14:46:40 can the same tree be used for cross and native builds? Jul 19 14:46:58 zecke: ping Jul 19 14:47:16 If i'm working on a piece of sw and want to build for arm and x86.... Jul 19 14:47:20 thanks reenoo_ Jul 19 14:47:59 is it better to work in seperate trees with seperate local.conf or is there a way to do it in the same tree Jul 19 14:48:19 to share the toolchain foo-native packages Jul 19 14:48:31 include multimachine.inc and you should be fine Jul 19 14:50:51 wow, that looks easy, a three line script which tweaks the structure :) Jul 19 14:55:22 epronk: I'm still gratefull to kergoth for coming up with that :) Jul 19 14:55:33 koen: if my local.conf says arm and I want to build for example libpng for x86 how do I specify it? Jul 19 14:55:42 good nite guys Jul 19 14:55:46 'night zecke Jul 19 14:55:52 bye Jul 19 14:56:10 epronk: I have the MACHINE parameter in auto.conf and switch that during builds Jul 19 14:57:47 you do a include auto.conf in local.conf? Jul 19 14:58:00 that get's done automatically Jul 19 14:58:09 undocumented feature :) Jul 19 14:58:32 so, out of the box it looks for a auto.conf? Jul 19 14:58:54 and this is the way to tweak things temporary Jul 19 14:59:11 yes Jul 19 14:59:12 i put my local.conf under version control Jul 19 14:59:15 ah Jul 19 15:00:26 I do something like: Jul 19 15:00:26 for i in h3600 h3900 ipaq-pxa-2.6 openzaurus-pxa-2.6 simpad collie corgi poodle ; do Jul 19 15:00:26 echo "MACHINE = \"$i\"" > conf/auto.conf Jul 19 15:00:27 echo "bitbaking $i" Jul 19 15:00:34 bitbake bootstrap-image ; bitbake gpe-image ; bitbake opie-image ; done Jul 19 15:00:54 which are incidentally all ARM machines, but you get the idea Jul 19 15:03:26 wow Water Polo looks exciting Jul 19 15:03:36 Kajap Polo I mean Jul 19 15:03:43 Kajak/Qanu.. whatever Jul 19 15:03:47 i'm off, see ya Jul 19 15:03:54 thanks for your help people Jul 19 15:04:00 take care Jul 19 15:04:04 zecke: see you in less then a week :P Jul 19 15:04:05 don't get stolen Jul 19 15:04:13 stolen? Jul 19 15:04:31 Stolen and sold into slavery :} Jul 19 15:04:32 zecke: kanopolo? Jul 19 15:04:42 Europe is just another 3rd world country Jul 19 15:04:44 koen: thanks Jul 19 15:04:56 koen: yeah whatever, it looks cool Jul 19 15:05:07 zecke: okay then :P Jul 19 15:05:10 we all should have a get together and play that Jul 19 15:05:30 the kanopolo teams trains 200m from my house Jul 19 15:07:44 ~bon appetit Jul 19 15:07:45 i heard bon appetit is smacznego. Guten Appetit. Eet Smakelijk. God Appetitt Jul 19 15:13:28 xya Jul 19 15:13:52 Appe Tit as my cousins used to say Jul 19 15:27:13 03koen 07org.oe.dev * r4ebdb5bb... 10/packages/ (gpsdrive/gpsdrive_2.10pre3.bb meta/gpe-image.bb): Add new gpsdrive prerelease, disable it for now till I finish testing it Jul 19 16:28:05 03jbowler 07org.oe.nslu2-linux * r80a8130b... 10/conf/distro/openslug.conf: Jul 19 16:28:05 OpenSlug-2.3-beta Jul 19 16:28:05 Tagged revision to capture OpenSlug-2.3-beta Jul 19 16:28:08 03jbowler 07org.oe.nslu2-linux * rf7ca9da6... 10/conf/distro/openslug.conf: The head is OpenSlug-2.4-beta to avoid confusion. Jul 19 16:39:48 'night all Jul 19 16:44:18 who was looking for lat/long of OE developers the other day? Jul 19 16:44:38 it was koen Jul 19 16:46:06 I'lll email him Jul 19 16:53:35 Hi! ... does oe have the ability to generate a 'relocatable' compiler toolchain so if I build an image and provide a toolchain other people could compile stuff for themselves without having to do the whole oe thing? Jul 19 16:54:31 I can clarify and provide an example if it would make my question easier to answer Jul 19 16:59:06 martin__: short answer is no - not really Jul 19 16:59:15 at least if you're talking about a cross compiler toolchain Jul 19 16:59:20 yea Jul 19 16:59:23 that is what I am talking about Jul 19 16:59:32 of course, OE does a fine job of packaging up the target toolchain Jul 19 16:59:52 but taht will only run on the arch of the device, right? Jul 19 17:01:49 the thing is... for the lab I work in, people often like to just compile stuff and have the pxa devices mount the workstations and run their software that way Jul 19 17:02:25 and the arm tool chains I know about do not yet support the floating point library stuff Jul 19 17:02:59 (or maybe I am just confused) Jul 19 17:05:27 allo? Jul 19 17:10:25 hmmm... I guess I have to figure out how to build a toolchain then Jul 19 17:12:47 what would be cool is if or could package the cross compiling toolchain and populate it with any lirbraries from the images it builds... so then I would have to do things like add libcurl and what not (like I currently do for the existing arm toolchains) Jul 19 17:23:28 damn old autoconf recipes....now it's tcl.... Jul 19 17:23:34 ::sigh:: Jul 19 17:33:13 ah...CVS has a recent configure.in Jul 19 20:57:34 * chouimat is back. Jul 19 21:45:43 "Some people are like Slinkies. Not really good for anything, but they still bring a smile to your face when you push them down a flight of stairs." Jul 19 21:45:57 kergoth: hey Jul 19 21:46:05 hey Jul 19 21:48:51 night Jul 19 23:10:43 CP|Asleep: bitbake meta-sdk will give you a toolchain ipk for your host Jul 19 23:25:38 hi all Jul 19 23:26:59 hey koobla Jul 19 23:28:57 morning koen Jul 19 23:30:38 hey hooooo ! Jul 19 23:31:46 hi alan_ Jul 19 23:32:45 yop yop koobla. Jul 19 23:33:37 i wonder why i made a build 4 days ago, made another one yesterday, and the first one works better than the second one on my zaurus... **** ENDING LOGGING AT Tue Jul 19 23:59:56 2005