**** BEGIN LOGGING AT Fri May 26 09:59:57 2006
May 26 10:03:47 <Bernardo> good morning
May 26 10:04:05 <hrw|work> hi Bernardo 
May 26 10:04:45 <koen> hey Bernardo 
May 26 10:05:05 <CIA-9> 03lenehan 07org.oe.dev * rf838bfb7... 10/packages/arpwatch/arpwatch_2.1a13.bb: arpwatch 2.1a13: Add fakeroot-native to DEPENDS since it's required to build. Closes #946.
May 26 10:05:10 <CIA-9> 03lenehan 07org.oe.dev * ra28eb033... 10/site/ (sh4-linux sh4-linux-uclibc): 
May 26 10:05:10 <CIA-9> site/sh4: Update the sh4 site files. This contains all of the changes over
May 26 10:05:10 <CIA-9> collected over the past 6 months for sh4 and gets a large amount of the non
May 26 10:05:10 <CIA-9> UI related components building.
May 26 10:11:32 <zecke> rwhitby: I have two 'issues' with NSLu2 :)
May 26 10:12:42 <zecke> rwhitby: first could you add the patch from http://bugs.openembedded.org/attachment.cgi?id=773 and test it?
May 26 10:13:02 <zecke> rwhitby: and the second machine/include/ixp4xx.conf resets OVERRIDES
May 26 10:13:06 <zecke> rwhitby: removing fail-fast
May 26 10:23:29 <koen> zecke: is the svg done yet?
May 26 10:24:06 <zecke> koen: I'm a moron ;)
May 26 10:24:33 <zecke> koen: I forgot to let it run in a screen...
May 26 10:25:17 <rwhitby> zecke: is that bitbake patch in svn?
May 26 10:25:40 <zecke> rwhitby: it is a candidate for backport
May 26 10:25:45 <rwhitby> and for the second one, are you able to commit a fix?
May 26 10:25:49 <zecke> rwhitby: it will be in svn in a day or two
May 26 10:26:25 <zecke> rwhitby: no 'proper' one. You want to add ixp4xx in the middle. I will commit a fix though
May 26 10:26:53 <rwhitby> thx - I expect you'd be far more successful with a fix than I would be.
May 26 10:27:35 <zecke> rwhitby: hehe, I will copy and paste the line from bitbake.conf again :}
May 26 10:28:35 <rwhitby> zecke: I won't have time in the next few days to test that bitbake patch unfortunately.  A busy weekend with no time for hacking.
May 26 10:29:53 <zecke> rwhitby: that is fine
May 26 10:30:35 * koen searches for some breakfast
May 26 10:37:47 <koen> like|zzz: I sent a mail to the person who reported #90 asking for a solution
May 26 10:39:23 <cyrilRomain> koen: cool
May 26 10:41:08 <cyrilRomain> koen: I tested different OE tarballs to determine when this bug arised
May 26 10:41:56 <cyrilRomain> koen: what is strange is that the glibc2.4 compiled without problem 1 week ago, and it was with the same glibc tarball ....
May 26 10:43:03 <cyrilRomain> koen: with OE-20060518.tar.bz2, glibc2.4 do_stage does not loop, whereas it loops with OE-20060519.tar.bz2 
May 26 10:44:48 <cyrilRomain> koen: but I don't find striking changes... PARALLEL_MAKE = "" was added to glibc_2.4.bb but removed on 20-5, so should not be that ....
May 26 10:59:08 <koen> cyrilRomain|food: http://ewi546.ewi.utwente.nl/OE/OE-20060518-20060519.diff.bz2
May 26 10:59:10 <koen> that's the diff
May 26 11:02:38 <RP> morning all
May 26 11:02:43 <koen> hey RP 
May 26 11:02:50 <RP> jacques: Did you want me?
May 26 11:03:07 <koen> cyrilRomain|food: I suspect the -isystem change makes glibc loop
May 26 11:03:20 <hrw|work> hi RP 
May 26 11:03:41 <hrw|work> RP: nodocs.patch in qemu-native is why jacques want you
May 26 11:04:14 <RP> hrw|work: Did qemu add a way to disable the document generation?
May 26 11:04:50 <RP> zecke: Do you know anything about the manifest code in oe/bitbake?
May 26 11:05:38 <zecke> RP: no
May 26 11:05:54 <zecke> RP: you emit the manifest task in base.bbclass?
May 26 11:06:21 <RP> zecke: One of my packages has a manifest file in it which causes a few issues
May 26 11:06:28 <hrw|work> RP: dont know - .oz354x has qemu 0.7.1
May 26 11:06:38 <RP> zecke: It also highlights that the manifest code is suffering bitrot
May 26 11:07:01 <RP> zecke: Since nobody is using it, why is it there? :)
May 26 11:07:04 <zecke> RP: it does, what is the MANIFEST used for (I only know it from python)
May 26 11:07:15 <zecke> RP: dunno, ask kergoth
May 26 11:08:12 <RP> zecke: I suspect it can become its own class at the very least
May 26 11:08:33 <zecke> RP: this is almost for sure
May 26 11:08:40 <RP> kergoth: ping :)
May 26 11:09:09 <woglinde> hi koen
May 26 11:09:14 <koen> hey woglinde 
May 26 11:09:23 <woglinde> made bugreport for msp430
May 26 11:09:35 <woglinde> 1045
May 26 11:09:56 <RP> hrw|work: I have some patches for qemu so I'll refresh that patch when I deal with that
May 26 11:10:27 <woglinde> hm new java update
May 26 11:12:55 <_guillermo> woglinde: what java update?
May 26 11:13:05 <woglinde> 1.5.0_uodate7
May 26 11:13:08 <woglinde> ups padte7
May 26 11:13:26 <_guillermo> amm, I though the update would be now it's open source
May 26 11:13:26 <woglinde> from sun
May 26 11:13:36 <woglinde> not yet
May 26 11:13:48 <woglinde> hihi
May 26 11:13:50 <woglinde> 6367204 hotspot garbage_collector CMS: Poor garbage collection performance
May 26 11:14:59 <_guillermo> zecke: the pcmciautils problem is still on oz354x, should I report a bug or will it be merged with .dev?
May 26 11:15:25 <woglinde> guillermo I would file a bug
May 26 11:15:34 <woglinde> which than should be closed
May 26 11:15:45 <hrw|work> _guillermo: simple copy from .dev is enough?
May 26 11:15:48 <zecke> _guillermo: reopen the bug :)
May 26 11:16:37 <_guillermo> hrw|work: I guess, it was just changing a line from "/usr/sbin/ln" to "ln" basically
May 26 11:17:02 <hrw|work> ok willpush
May 26 11:17:25 <zecke> RP: I think we can almost savely remove that
May 26 11:17:46 <RP> zecke: From OE, bitbake or both? :)
May 26 11:18:29 <zecke> RP: OE at first
May 26 11:19:54 <mickeyl> the manifest stuff was a cool idea
May 26 11:20:00 <pb__> hail mickeyl
May 26 11:20:04 <woglinde> hi mickeyl
May 26 11:20:07 <mickeyl> the idea was to automatically get the files that are necessary to make do_install
May 26 11:20:16 <mickeyl> good morning pb, woglinde, et.al.
May 26 11:20:19 <koen> hey mickeyl 
May 26 11:20:23 <RP> hi mickeyl 
May 26 11:20:34 <mickeyl> alas, i don't think we need it atm., so feel free to externalize it
May 26 11:20:40 <mickeyl> hey RP
May 26 11:20:51 <mickeyl> RP: i pointed Harald to your kernel for implementation of
May 26 11:20:51 <hrw|work> hi mickeyl 
May 26 11:20:53 <mickeyl> a) backlight
May 26 11:20:59 <mickeyl> b) LED
May 26 11:21:03 <mickeyl> c) hinge sensor
May 26 11:21:13 <mickeyl> I hope he'll see the greatness of your code
May 26 11:21:17 <mickeyl> *crossing fingers*
May 26 11:21:35 <RP> mickeyl: Those are all nicely established in mainline now so lets hope so :)
May 26 11:21:39 <koen> mickeyl: I had a neat idea for a video player: resize the window on hinge close 
May 26 11:21:55 <mickeyl> koen: heh, cool idea,
May 26 11:22:21 <RP> mickeyl: I did agree to having the keyboard switch definitions redefined but the good news is we stay backwards compatible this time
May 26 11:22:33 <mickeyl> RP: ah good.
May 26 11:22:44 <mickeyl> RP: http://www.vanille.de/temp/a780
May 26 11:22:57 <mickeyl> slowly, but progressing :)
May 26 11:23:07 <RP> mickeyl: It also means opie can just listen for the signal under 2.6 and always use it to rotate the display etc.
May 26 11:23:48 <RP> mickeyl: http://www.rpsys.net/openzaurus/patches/input_switch_rename-r0.patch
May 26 11:23:50 <zecke> mickeyl: do you have your phone yet?
May 26 11:23:59 <mickeyl> zecke: yes! :D
May 26 11:24:06 <mickeyl> zecke: the photos are taken by me
May 26 11:24:27 <cyrilRomain> koen: well, I'll try with -isystem removed. Is it safe ?
May 26 11:24:31 <RP> mickeyl: With this manifest code, I still don't understand what its meant to do. Was this something for precompiled binaries?
May 26 11:24:35 <mickeyl> RP: ah ok, that's ok
May 26 11:25:09 <hrw|work> _guillermo: pushed
May 26 11:25:15 <koen> cyrilRomain: I'm building as we speak :)
May 26 11:25:19 <mickeyl> RP: iirc the idea was to have a manifest-generating mode in which we run do_install to generate manifest for a package and then on subsequent runs we use this manifest to decide what files to put where
May 26 11:25:21 <RP> mickeyl: It should make 2.6 device abstraction a bit easier
May 26 11:25:30 <_guillermo> mickeyl: the pictures are promising, what are the biggest difficulties now with the a780?
May 26 11:25:33 <cyrilRomain> koen: ok :)
May 26 11:26:01 <mickeyl> _guillermo: biggest difficulties are kernel wise atm. problems in early init code, AP/BP handshake, and of course... accessing the BP at all
May 26 11:26:20 <mickeyl> (BP=Baseband Processor, the one that does all the phone stuff)
May 26 11:26:24 <mickeyl> (AP=Application Processor)
May 26 11:26:26 <CIA-9> 03hrw 07org.oe.oz354x * r4e02b929... 10/packages/pcmciautils/ (3 files in 2 dirs): pcmciautils: got fixes from .dev - should not segfault now
May 26 11:26:52 <_guillermo> mickeyl: how much code and info is motorola releasing?
May 26 11:26:53 <RP> mickeyl: ok. I'll look at moving it into a manifest class. At least I have some idea what it was meant to do now, thanks :)
May 26 11:27:05 <mickeyl> _guillermo: unbelievably much. we have nearly everything we asked for
May 26 11:27:11 <mickeyl> RP: :)
May 26 11:27:24 <koen> and motorola is opensourcing stuff as well
May 26 11:27:43 <zecke> koen: hehe
May 26 11:27:57 <mickeyl> RP: i think the idea of a universal device daemon will be realized sooner or later
May 26 11:27:59 <mickeyl> RP: http://bugs.openembedded.org/show_bug.cgi?id=347
May 26 11:28:02 <_guillermo> mickeyl: that's great, I don't know if it's just me but I feel a strong change in embedded industry about open source
May 26 11:28:12 <mickeyl> RP: I'm fed up of doing anything twice and more :)
May 26 11:28:45 <mickeyl> _guillermo: It's not you. There is a change. It's slowly, but I think it's unstoppable
May 26 11:29:07 <ade|desk> ah can someone tell me about overrides, please
May 26 11:29:19 <cyrilRomain> koen: any answer from Boris Kolpackov ?
May 26 11:29:41 <koen> cyrilRomain: not yet
May 26 11:29:50 <likewise> cyrilRomain: hi there
May 26 11:29:57 <koen> hey likewise 
May 26 11:29:59 <likewise> koen: hi
May 26 11:30:03 <cyrilRomain> hey likewise 
May 26 11:30:11 <_guillermo> anyone remember what package was giving problems with akita and the touchpad in .dev? I want to do some testing
May 26 11:30:24 <likewise> cyrilRomain: I see you found the right diff, right?
May 26 11:30:30 <ade|desk> there is an EXTRA_OECONF_uclibc = "someoption" but it seems to be overlooked ... is that the correct way to have overrides for uclibc builds?
May 26 11:30:37 <cyrilRomain> likewise: yep
May 26 11:30:45 <zecke> koen: wonderful :}
May 26 11:31:17 <likewise> koen: probably the -isystem affects glibc (which maybe is broken then). I will start a build without -isystem, but maybe with fail_fast enabled.
May 26 11:31:39 <cyrilRomain> likewise: but not the thing patch that makes the loop. Koen is trying without the -isystem line in bitbake.conf
May 26 11:32:12 <koen> the build can't fail now
May 26 11:32:24 <koen> I'm wearing my o-hand shirt :)
May 26 11:32:28 <likewise> cyrilRomain: OK, koen is already on it then. "the thing patch"??
May 26 11:32:32 <RP> mickeyl: I can see a role for zaurusd renamed as devmand for this
May 26 11:32:35 <likewise> :-)
May 26 11:33:01 <_guillermo> how come I have my oz354x tree updated but I see only openzaurus rc2 and not rc4?
May 26 11:34:02 <mickeyl> RP: righto. I even see that I could write this daemon in C++ as no one will care unless it exports everything over a dbus interface :)
May 26 11:35:00 <cyrilRomain> likewise: yes the 'thing patch' or 'the modification that cause the loop'
May 26 11:35:02 <koen> speaking of shirts...
May 26 11:35:11 * koen mails the OE logo to the printshop
May 26 11:35:12 <mickeyl> mallum promised me a o-hand shrit
May 26 11:35:18 <mickeyl> i still don't have one :)
May 26 11:35:19 * cyrilRomain should really re-open some english course book
May 26 11:35:29 <mickeyl> koen: yay! :)
May 26 11:35:34 <RP> mickeyl: I was wondering about puting a lot of the device specific config into zaurusd. It could easily absorb things like the current ts-conf for example
May 26 11:36:07 <mickeyl> RP: exactly. a single point of device knowledge accesable from all GUI environments and also console stuff would be much appreciated
May 26 11:36:26 <mickeyl> RP: I really woudl like to remove most deviceisms from Opie and GPE and E
May 26 11:36:32 <RP> mickeyl: Yes. I'm just unsure how that would do down with other OE devs
May 26 11:37:08 <RP> mickeyl: I can see some people saying that they don't want to have to install devmand, One aim of devmand is that is will optimise itself to the smallest possible size though
May 26 11:37:54 <koen> mickeyl: just to get a quote and hints on how to adjust the logo
May 26 11:38:12 <RP> mickeyl: I will remind mallum ;-)
May 26 11:38:16 <mickeyl> RP: thanks :)
May 26 11:39:20 <mickeyl> RP: the major thing concerning me is latency. deviceisms in the core have the advantage of being fast. if it will take 5 seconds to sense a rotation from hardware to opie that would be unacceptable for me
May 26 11:39:55 <koen> I think the best way is that someone add dbus support to zaurusd/devmand and people can discuss it from there
May 26 11:40:02 <koen> instead of getting stuck at the design
May 26 11:40:02 <mickeyl> righto
May 26 11:40:19 <mickeyl> well, a bit of design can't hurt :)
May 26 11:40:27 <RP> At present there is code for an switchevd in zaurusd. The idea was always to interface that to dbus
May 26 11:40:30 <mickeyl> at least for the object ree / path
May 26 11:40:36 <mickeyl> RP: excellent.
May 26 11:40:39 <RP> There should be little latency from  switchevd
May 26 11:41:02 <RP> switchevd is only built/installed for devices that say they need it
May 26 11:45:34 * chouimat is away: long and  boring test .... 4 hours :(
May 26 11:48:14 <Frg> what is the prefered way to add some missing -llibrary to makefile's LIBS=...   withought changing package sources?
May 26 11:48:47 <mickeyl> depends on the buildsystem
May 26 11:48:52 <mickeyl> in your case:
May 26 11:49:20 <mickeyl> EXTRA_QMAKEVARS_POST += "LIBS+=-lopietooth2"
May 26 11:49:22 <mickeyl> i guess
May 26 11:49:39 <mickeyl> although i'm still questioning why that is necessary on your system and not on mine
May 26 11:49:54 <Frg> im missing -lopiebluez in many packages
May 26 11:50:13 <mickeyl> I'd rather inquiry upstream than to fix that in OE
May 26 11:50:21 <mickeyl> please write a mail to opie-devel@handhelds.org
May 26 11:50:21 <Frg> else it shows unreferenced symbols in libqpe
May 26 11:50:48 <hrw|work> atleast opie upstream react fast
May 26 11:51:12 <Frg> what should i put there
May 26 11:51:47 <mickeyl> find out which packages need -lopiebluez and ask them if they forgot this dependency or if it's intentional
May 26 11:52:00 <mickeyl> might also add that it doesn't occur on all systems...
May 26 11:52:17 <likewise> all: How can I enable the fail-fast feature (which is an override)? I have no clue where to make it condition on.
May 26 11:52:24 <Frg> ok.. its already 4th package
May 26 11:53:08 <cyrilRomain> likewise: SRC_URI_append_fail-fast = " file://zecke-no-host-includes.patch;patch=1 " ?
May 26 11:53:56 <likewise> cyrilRomain: OK, yes, that seems quite logical :-)
May 26 11:54:15 <cyrilRomain> and I think you need OVERRIDES = "local:${MACHINE}:${DISTRO}:${TARGET_OS}:${TARGET_ARCH}:build-${BUILD_OS}:fail-fast" in bitbake.conf
May 26 11:58:51 <koen> NOTE: package glibc-2.4: completed
May 26 11:58:59 <koen> yup, the isystem break glibc 2.4
May 26 11:58:59 <cyrilRomain> koen: Woohoo !
May 26 11:59:31 <cyrilRomain> koen: well done koen :)
May 26 11:59:47 <koen> cyrilRomain: you tracked it down to a certain date
May 26 12:01:51 <zecke> koen: hmm
May 26 12:02:04 <koen> yes, glibc is broken
May 26 12:04:12 <RP> hrw|work: PR went backwards to pcmciautils in the oz345x branch
May 26 12:05:50 <hrw|work> ups
May 26 12:06:35 <likewise> cyrilRomain: But where is fail-fast being set then. I mean the SRC_URI_append_* is conditional on *.
May 26 12:07:14 <hrw|work> will push r4
May 26 12:07:22 <likewise> koen: can we further track this down using zecke's patch that detects host includes?
May 26 12:07:39 <koen> likewise: I have no idea
May 26 12:07:41 <zecke> likewise: OVERRIDES
May 26 12:07:57 <likewise> OVERRIDES = "fail-fast" ?
May 26 12:08:26 <zecke> likewise: doing that alone would be stupid, but we apply this patch conditionally
May 26 12:08:34 <zecke> likewise: conditionally to OVERRIDES (see bitbake.conf)
May 26 12:09:03 <likewise> zecke: Yes, I am trying to understand the OVERRIDE system from bitbake - http://bitbake.berlios.de/manual/ch02.html#id2474646
May 26 12:09:34 <CIA-9> 03hrw 07org.oe.oz354x * r3eb87dad... 10/packages/pcmciautils/pcmciautils_013.bb: pcmciautils: bump PR
May 26 12:09:44 <zecke> likewise: everything behind a '_' is considered a override
May 26 12:10:00 <zecke> likewise: if you put it in OVERRIDES and call update_data it will be 'expanded' correctly
May 26 12:10:08 <sws> isnt oz354x supported on poodle? i build a gpe-image and it hangs while touch screen calibration
May 26 12:10:31 <koen> sws: #openzaurus ?
May 26 12:10:38 <likewise> zecke: Yes, in OVERRIDES you specify the conditional properties, and after _ you specify the *value* of the property? as in MACHINE, arm.
May 26 12:11:19 <zecke> likewise: SRC_URI_openzaurus
May 26 12:11:28 <zecke> likewise: and you have ${DISTRO} in OVERRIDES
May 26 12:13:39 <Frg> figured out that those symbols are defined in more libs than only that one mentioned, namely:
May 26 12:13:42 <Frg> libopiebluez2.so libopiemm2.so libopienet2.so libopiepim2.so libopietooth1.so libopieui2.so libqtaux2.so
May 26 12:14:02 <likewise> zecke: clear, I was thinking too hard there.  :-)
May 26 12:14:05 <Frg> so maybe linking against qtaux or opieui2 would be wiser
May 26 12:14:40 <zecke> Frg: compare the compile options from Qt2, libqpe, libopie
May 26 12:14:50 <zecke> (e.g. -fno-exceptions -fno-rtti)
May 26 12:20:22 <likewise> rm_work is broken in the sense that it sometimes stops the build, without nasty side effects other than that the dir is not completely rm-ed.
May 26 12:20:57 <CIA-9> 03koen 07org.oe.packaged-staging * r18e997ef... 10/packages/gcc/ (8 files in 2 dirs): 
May 26 12:20:57 <CIA-9> gcc: add 4.1.1 from .dev
May 26 12:20:57 <CIA-9> * revision ed6d863c36fd1d9d7ab0dba3d700e0f1ccf122be
May 26 12:20:57 <CIA-9> * revision dcfd1f3cb8ffcc49243af19ba6eaf85d4ef9c04a
May 26 12:20:57 <CIA-9> * revision 48950da6e0a596e94a112da02ea2bd3f5d6c6782
May 26 12:21:02 <CIA-9> 03koen 07org.oe.packaged-staging * ra52a49ac... 10/conf/distro/ (3 files): Add missing OZ bits
May 26 12:21:06 <CIA-9> 03koen 07org.oe.packaged-staging * rbead2b7c... 10/conf/distro/angstrom-2006.9.conf: 
May 26 12:21:06 <CIA-9> angstrom 2006.9: use gcc4.
May 26 12:21:06 <CIA-9> * revision 2d6d863c36fd1d9d7ab0dba3d700e0f1ccf122be
May 26 12:21:10 <CIA-9> 03koen 07org.oe.packaged-staging * r4e82ed47... 10/conf/bitbake.conf: bitbake.conf: turn of -isystem till glibc gets fixed
May 26 12:23:28 <zecke> koen: why not set a different CFLAGS in glibc?
May 26 12:23:49 <koen> zecke: this was easier
May 26 12:25:42 <koen> I'm all set now, time to start a new .bbclass
May 26 12:26:16 <zecke> koen: wanna talk?
May 26 12:33:17 <hrw|work> source fetching still in progress...
May 26 12:43:41 <CIA-9> 03koen 07org.oe.packaged-staging * r29a1e616... 10/packages/ncurses/ (9 files in 2 dirs): Add ncurses
May 26 12:43:46 <CIA-9> 03koen 07org.oe.packaged-staging * rbab1aab2... 10/packages/nano/ (nano-1.3.9/glib.m4 nano_1.3.9.bb): Add nano
May 26 12:43:50 <CIA-9> 03mickeyl 07org.oe.dev * rb9e1ee8d... 10/packages/tslib/ (tslib/a780/tslib.sh tslib/e680/tslib.sh tslib_cvs.bb): tslib cvs: add support for the Motorola E680 and A780
May 26 12:43:54 <CIA-9> 03mickeyl 07org.oe.dev * r10f03b89... 10/classes/sanity.bbclass: sanity.bbclass: improve error message, check for cvs and svn
May 26 12:43:56 <hrw|work> koen: bitbake nano?
May 26 12:43:58 <CIA-9> 03mickeyl 07org.oe.dev * r13b957f3... 10/packages/qte/ (7 files): qte: clean up, remove double entries, add flags for Motorola A780 and E680 GSM phones
May 26 12:44:02 <CIA-9> 03mickeyl 07org.oe.dev * r54290256... 10/packages/libusb/ (libusb_0.1.8.bb libusb-native_0.1.12.bb libusb_0.1.12.bb): add libusb[-native] 0.1.12
May 26 12:44:04 <koen> hrw|work: yes :)
May 26 12:44:07 <CIA-9> 03mickeyl 07org.oe.dev * rd0e9c0f4... 10/packages/base-files/ (3 files in 3 dirs): base-files: add tmpfs to stock fstab
May 26 12:44:11 <CIA-9> 03mickeyl 07org.oe.dev * r6c892a12... 10/packages/linux/ (4 files in 2 dirs): linux-ezx: defconfig updates, add known good one as backup
May 26 12:44:35 <hrw|work> koen: nano is most known package in OE probably ;)
May 26 12:44:53 <koen> yeah and it doesn't have too many deps
May 26 12:45:14 <hrw|work> only libc + ncurses iirc
May 26 12:45:18 <zecke> koen: ALT+ENTER
May 26 12:45:20 <koen> indeed
May 26 12:45:27 <koen> zecke: wrong window ;)
May 26 12:45:48 <likewise> I'm testing this fix for rm_work:
May 26 12:46:02 <likewise>         if [ `basename ${S}` == $dir ]; then
May 26 12:46:04 <likewise>             rm -rf $dir/.*   <----- Added, as dirs such as .pc/ remained
May 26 12:46:05 <likewise>             rm -rf $dir/*
May 26 12:46:27 <koen> .[a-z]* ?
May 26 12:46:27 <hrw|work> likewise: any fixes welcome - I wrote it years ago and rarely used
May 26 12:46:48 <koen> wouldn't .* match '..' ?
May 26 12:47:06 <likewise> hrw|work: well, I still find it strange why rm_work stops the build sometimes, even with this.
May 26 12:47:35 <likewise> koen: then a recursive delete would have deleted my whole dis*&(*@&@)(#&#)(
May 26 12:47:50 <koen> it happened to me once
May 26 12:48:09 <koen> (while being root)
May 26 12:48:29 <cyrilRomain> koen: should angstrom people switch to the packaged-staging branch ?
May 26 12:48:47 <koen> cyrilRomain: no
May 26 12:48:47 <hrw|work> cyrilRomain: noone should switch to it rather
May 26 12:49:44 <koen> right
May 26 12:49:51 <koen> unless you only need nano ;)
May 26 12:50:14 <cyrilRomain> koen: ok, just wondering if bitbake.conf changes are pushed into .dev too ?
May 26 12:50:45 <hrw|work> cyrilRomain: packaged-staging branch is koen's SoC branch
May 26 12:50:55 <likewise> [leon@dhcppc7 gdk-pixbuf]$ rm -rfv bla/.*
May 26 12:50:57 <likewise> rm: cannot remove `.' or `..'
May 26 12:50:58 <likewise> But you are right, probably safer to do .[a-z]*
May 26 12:51:24 <cyrilRomain> hrw|work: thx for the info
May 26 12:51:27 <CIA-9> 03hrw 07org.oe.oz354x * ra16d1cde... 10/packages/xmame/ (4 files in 2 dirs): xmame: added xmame-tiny from .oz354fam083, added fix from #182 - now it works
May 26 12:52:23 <cyrilRomain> likewise: yes but if the file start with a number ? hm... maybe .[^.]* ?
May 26 12:53:12 <likewise> cyrilRomain: much better indeed
May 26 12:53:42 <cyrilRomain> likewise: or even better .[^.]+ ?
May 26 12:54:27 * cyrilRomain think .[^.]* would probably complain about removing '.'
May 26 13:12:07 <cyrilRomain> any kernel hacker around ? linux-openzaurus-2.6.16-r33 does not build here. http://oe.pastebin.com/738208
May 26 13:12:51 <hrw|work> hmm.. not tested this with eabi... with eabi it was built
May 26 13:12:53 <hrw|work> hmm.. not tested this with eabi... with abi it was built
May 26 13:14:52 <cyrilRomain> i'm building with eabi, note that with eabi the 2.6.16-r25 built correctly
May 26 13:15:28 <RP> cyrilRomain: hrw added squashfs since then. It appears to expose a bug in the eabi toolchain looking at that log
May 26 13:15:52 <koen> or in the kernel :)
May 26 13:16:07 <RP> cyrilRomain: If you don't need squashfs, disable that in the defconfig as a workaround. It'd be nice to work out what causes the toolchain to break though
May 26 13:16:12 <cyrilRomain> RP: ok. I looked into the inode.c file but no consistency (for my eyes)
May 26 13:16:38 <RP> koen: That error looks more like a compiler issue although I could be wrong
May 26 13:17:04 <koen> I'm neither a toolchain export nor a kernel coder
May 26 13:17:09 <cyrilRomain> RP: well I'd say that too, but I got it on two build host :-/
May 26 13:17:43 <RP> cyrilRomain: The build host shouldn't change the generated code so thats not surprising
May 26 13:18:04 <cyrilRomain> RP: right
May 26 13:25:50 <CIA-9> 03koen 07org.oe.packaged-staging * rc47a5fb3... 10/classes/packaged-staging.bbclass: classes/packaged-staging.bbclass: add skeleton
May 26 13:26:56 <cyrilRomain> RP: thanks for the defconfig workaround :) Should I file a bug for this build issue ?
May 26 13:32:44 <pb__> cyrilRomain: that's a compiler bug.  you'd need to run it under the debugger to find out where the error is occurring.
May 26 13:35:33 <cyrilRomain> pb__: well, gdb <the command that build the kernel> ?
May 26 13:35:47 <cyrilRomain> pb__: i'm not a debugger expert bug will try
May 26 13:39:34 <likewise> cyrilRomain: gdb --args <the command with args like you normally put on the command line>
May 26 13:40:10 <cyrilRomain> likewise: thanks
May 26 13:43:45 <cyrilRomain> ouch, dropbear issue now http://oe.pastebin.com/739259. well let's debug the kernel issue then
May 26 13:45:05 <zecke> http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=poky <- we now have a poky tree in the tinderbox
May 26 13:45:08 <zecke> *hint*
May 26 13:51:48 <CosmicPenguin> OLPC developer boards should be coming available soon - I know they are using a Fedora image, but it probably wouldn't hurt if OE shadowed them in some way
May 26 13:52:08 <CosmicPenguin> I'll do my part and keep the kernel and other such things up to date
May 26 13:52:33 <zecke> CosmicPenguin: who can get OLPC boards?
May 26 13:52:46 <CosmicPenguin> Its not really clear at this point
May 26 13:52:51 <hrw|work> cyrilRomain: hmm. it does not have zlib in DEPENDS?
May 26 13:52:53 <CosmicPenguin> http://wiki.laptop.org/index.php/Notes_on_using_the_OLPC_developer_boards
May 26 13:53:18 <CosmicPenguin> zecke: but I'm sure that somebody here has enough respect within hh.o to convince Jim Gettys to give them one
May 26 13:54:01 <cyrilRomain> hrw|work: it has
May 26 13:54:26 <zecke> CosmicPenguin: I think pb_ is the best candidate
May 26 13:54:50 <cyrilRomain> hrw|work: probably a missing include dir...
May 26 13:55:53 <CosmicPenguin> zecke: I agree - he would be good
May 26 13:56:43 <CosmicPenguin> From the website, it looks like Matthew Allum and Fabritio Sampsa are also mentioned in regards to the operating environment, so maybe o-handed can get one or two too
May 26 13:57:30 <zecke> hmm why is my nfsroot not working
May 26 14:03:58 <RP> CosmicPenguin: s/handed/hand/ ;-)
May 26 14:04:23 <zecke> hehe
May 26 14:05:04 <RP> zecke: poky tinderbox support is an interesting idea...
May 26 14:05:28 <zecke> RP: I hope to get off my ass to write a OE Tinderclient tool... (today)
May 26 14:05:48 <zecke> RP: the qemu target is what interests me most (running tests)
May 26 14:06:09 <cyrilRomain> I'm a bit confused about eabi, should I set TARGET_OS = "linux-eabi" (cf koen mail) or "linux-gnueabi" (cf linux-openzaurus.inc) ?
May 26 14:07:44 <RP> zecke: I'll try and release some qemu images as I think they might interest a few people
May 26 14:08:51 <zecke> my tinderclient now has auto evict...
May 26 14:09:01 <cyrilRomain> zecke: for testcase, did you know http://litmus.mozilla.org/ ?
May 26 14:09:56 <zecke> cyrilRomain: no
May 26 14:10:03 <cyrilRomain> it is more designed towards end-user tests
May 26 14:10:17 <NAbyss> Hmm, is there any page on the OE wiki with info on how to track down build problems?
May 26 14:10:38 <zecke> NAbyss: not yet, but as it is long lived it belongs into the manual
May 26 14:11:05 <zecke> NAbyss: 1. look at the log 2. ??? 3. attach patch to the bug tracker
May 26 14:11:40 <NAbyss> zecke: 2 is where I'm getting stuck.. know of any descriptions of how bitbake works out what to do, apart from RTFS?
May 26 14:12:05 <zecke> NAbyss: what did fail? which task?
May 26 14:12:39 <cyrilRomain> NAbyss: you can try 'bitbake -i' for an interactive shell, 'bitbake -v -D -D' for more trace info
May 26 14:13:10 <NAbyss> zecke: Building gpsd, do_compile... I'm getting cross compile badness with the INCLUDE_PATH
May 26 14:13:15 <NAbyss> cyrilRomain: Ah, thanks
May 26 14:13:34 <zecke> NAbyss: hmm update bb  data?
May 26 14:13:46 <NAbyss> zecke: I did a pull this morning
May 26 14:13:47 <cyrilRomain> NAbyss: more '-D' means more trace depth
May 26 14:14:04 <zecke> NAbyss: monotone update? as well? IIRC mickeyl fixed that
May 26 14:14:19 <NAbyss> zecke: Ah, I'll give it another go
May 26 14:15:16 <zecke> NAbyss: we have a special gcc patch now. if you have -I/usr/include when cross compiling it will complain :)
May 26 14:15:37 <NAbyss> zecke: Yeah, very wise :)
May 26 14:18:04 <cyrilRomain> To enable eabi, should I set TARGET_OS = "linux-eabi" (cf koen mail) or "linux-gnueabi" (cf linux-openzaurus.inc) in my local.conf ?
May 26 14:18:24 * cyrilRomain used "linux" *sigh* :-[
May 26 14:20:26 <CosmicPenguin> RP: heh - indeed
May 26 14:21:27 <CosmicPenguin> http://www.pixelbeat.org/fslint/
May 26 14:21:33 <CosmicPenguin> Hmm... I should run that on my rootfs some time... :)
May 26 14:27:35 <CIA-9> 03freyther 07org.oe.dev * r7d2025aa... 10/conf/machine/include/ixp4xx.conf: 
May 26 14:27:35 <CIA-9> conf/machine/include/ixp4xx.conf: Enable fail-fast as well
May 26 14:27:35 <CIA-9>  OVERRIDES got redefined in this configuration file. We add
May 26 14:27:35 <CIA-9>  fail-fast now as well.
May 26 14:27:38 <CIA-9> 03freyther 07org.oe.dev * r4df4d7b4... 10/classes/tinderclient.bbclass: 
May 26 14:27:38 <CIA-9> classes/tinderclient.bbclass: Retry sending, Clean package when failing
May 26 14:27:38 <CIA-9>  Retry to connect the tinderbox a couple of times. This is needed
May 26 14:27:41 <CIA-9>  on machines behind broken routers.
May 26 14:27:44 <CIA-9>  When a package is broken (PackageFailed/TaskFailed) we will clean
May 26 14:27:46 <CIA-9>  the package so we can retry the next time.
May 26 14:31:19 <lardman|work> bye chaps
**** ENDING LOGGING AT Fri May 26 14:40:27 2006

**** BEGIN LOGGING AT Fri May 26 17:02:11 2006
May 26 17:02:38 <danboid> but there is a w100 driver for mplayer isn't there? Does it just do hardware video resize and thats it?
May 26 17:02:57 <RP> It uses the overlay capability of the w100
May 26 17:03:18 <RP> There are other capabilities of the w100 we know about too like fills and blits
May 26 17:04:34 <danboid> I'd never heard of overlays before I got my Z a few months ago. Do PC gfx cards offer a similiar feature (but much faster and bigger of course)
May 26 17:05:06 <RP> Yes, I'd expect so
May 26 17:05:19 <zecke> danboid: www.directfb.org ;)
May 26 17:05:36 <hrw> hi
May 26 17:05:44 <danboid> hi hrw!
May 26 17:05:56 <hrw> I heard that w100 can do idct...
May 26 17:06:30 <danboid> hrw: hows sashz doing ;)
May 26 17:07:05 <RP> hrw: I've heard that too. Having looked into the drivers, I'm still not sure if its true or not
May 26 17:07:18 <zecke> hrw: we should ask for X2 discounts, as I want one as well
May 26 17:07:46 <hrw> ~lart quilt
May 26 17:07:46 * ibot moos at quilt
May 26 17:08:43 <hrw> how to reorder patch series with quilt in other way then 'quilt pop 99999', rm -r patches;quilt import.... ;quilt push 9999?
May 26 17:09:21 <RP> hrw: quilt pop -a == quilt pop 99999
May 26 17:09:41 <RP> hrw: quilt pop -a, edit series file, quilt push -a ?
May 26 17:09:51 <CosmicPenguin> yeah - thats the best way to do it
May 26 17:09:54 <hrw> RP: yes
May 26 17:10:00 <CosmicPenguin> Some times I wish there was a way to move patches in quilt
May 26 17:10:13 <CosmicPenguin> but I really wish there was a way to move patches in stg, since it doesn't have a series file like quilt does
May 26 17:10:18 <hrw> time to learn how to rebuild external modules in debian with make-kpkg
May 26 17:10:47 <zecke> hrw: isn't there a nice dialog based tool?
May 26 17:11:11 <hrw> zecke: ?
May 26 17:11:24 <CosmicPenguin> dialog based quilt tool?
May 26 17:11:25 <zecke> hrw: to compile external modules?
May 26 17:11:26 * CosmicPenguin pukes
May 26 17:11:28 <ade|desk> crap, atexit() error with busybox and gcc4.1.0
May 26 17:11:41 <zecke> hrw: dialog assistant or such
May 26 17:12:48 <hrw> module-assistant - tool to make module package creation easier
May 26 17:13:03 <hrw> thx xze
May 26 17:15:41 <danboid> zecke: thanks for the link! I haven't checked out directfb in a long time and it has progressed massively since I last checked on it by the looks of things! Do many GTK apps support directfb? I want ROX under directfb!
May 26 17:16:08 <danboid> (The ROX desktop)
May 26 17:16:36 <zecke> danboid: the next debian installer will use DirectFB
May 26 17:16:52 <zecke> danboid: but it is an example for the usage of hardware accelerated laeyers (overlays)
May 26 17:17:45 * hrw will not touch rox-filer
May 26 17:19:16 <danboid> wow! Another great reason I can't wait for etch- not as excited about etch as etch+1. Any chance of us seeing the Zaurus fully supported on the ARM version of etch?
May 26 17:19:39 <danboid> hrw: whats wrong with rox in yr opinion?
May 26 17:20:16 <hrw> danboid: AppDir idea
May 26 17:20:32 * RP -> back later
May 26 17:20:43 <danboid> bye rp
May 26 17:20:56 <hrw> danboid: I maintained rox-filer for debian for over year (iirc) - had two versions.. vanilla and -hrw without appdir
May 26 17:21:33 <danboid> he! I didn't know that! You must like rox-filer a bit then?
May 26 17:22:01 <hrw> no
May 26 17:22:12 <danboid> whats your favourite filer then?
May 26 17:22:16 <danboid> gui
May 26 17:22:24 <hrw> mc
May 26 17:22:25 <zecke> danboid: what is fully supported?
May 26 17:22:25 <zecke> danboid: if you are lucky you can fit dpkg in the flash of the Zaurus ;)
May 26 17:22:31 <hrw> then konqueror
May 26 17:22:42 <zecke> mc!
May 26 17:22:56 <danboid> i see oe loves mc
May 26 17:23:13 <danboid> i'm a mac- m A c
May 26 17:23:19 <danboid> ..donald
May 26 17:23:40 <zecke> danboid: mc, then finder+spotlight
May 26 17:23:45 <danboid> eMM Aaa Ceee with a big D :D
May 26 17:24:05 <hrw> zecke: you press mouse button on left or right side?
May 26 17:25:00 <zecke> hrw: middle?
May 26 17:25:17 <danboid> zecke: don't you think the new installer will work on a 64MB machine?
May 26 17:25:32 <ade|desk> right home time .. uclibc may have to wait while i get my MSc stuff done for tuesday .. doh
May 26 17:25:41 <zecke> danboid: it is more the question of available flash
May 26 17:25:44 <danboid> 64MB RAM
May 26 17:25:46 <hrw> danboid: dont you think that Z with 23M rootfs has a bit too less space?
May 26 17:25:46 <ade|desk> night all
May 26 17:25:52 <hrw> cu ade
May 26 17:26:29 <danboid> yes but couldn't it install just like oe does on the cxxxx
May 26 17:27:18 <danboid> i'm only talking about seeing c3xxx support in debian
May 26 17:27:31 <hrw> I do not have c3k
May 26 17:27:34 <hrw> yet
May 26 17:27:41 <hrw> ;D
May 26 17:28:56 <kergoth> debian is a beast.  hardly suitable as is for any embedded application
May 26 17:29:05 <kergoth> if that wasnt the case, oe might not exist in its current form
May 26 17:29:12 <hrw> OT: what brand of DVDR media you suggest? tomorrow friend will borrow/lend me dvdrw so I will leave ~40GB of disk
May 26 17:29:14 <kergoth> heh
May 26 17:29:27 <hrw> kergoth: CNAME wiki.openzaurus.org -> openzaurus.berlios.de
May 26 17:29:32 <kergoth> k
May 26 17:29:32 <danboid> thats shyte- sharp should issue all the main oe devs with all the models
May 26 17:29:57 <goxboxlive> zecke: I switched to branch oz354fam083 and now opie and gpe builds as they should. Why is that, why dis i have problem with opie-bluepin with the dev branch and not the fam branch?
May 26 17:30:26 <hrw> goxboxlive: oz354fam083 and oz354x branches was made to provide STABLE thing
May 26 17:30:33 <danboid> kergoth: point taken
May 26 17:30:40 <hrw> goxboxlive: dev is DEVelopment
May 26 17:31:12 <hrw> goxboxlive: .dev = unstable == experiments == you have to know what you do
May 26 17:31:13 <danboid> 'debian is a beast' :p
May 26 17:31:21 <zecke> goxboxlive: fixed SRCDATE ;)
May 26 17:31:37 <zecke> goxboxlive: you might have even more fun with oz354x
May 26 17:31:51 <goxboxlive> ok
May 26 17:32:17 <kergoth> danboid: you might find it amusing that in wookie's embedded debian presentation, he advises that anyone wanting something actually usable should go use scratchbox or oe.
May 26 17:32:26 <kergoth> hehe
May 26 17:32:44 <zecke> kergoth: hey
May 26 17:32:54 <goxboxlive> so maybe there wasent any problem with my setup. Was it opie-bluepin that where broken?
May 26 17:33:07 <kergoth> zecke: hey
May 26 17:33:35 * kergoth is disappointed and somewhat annoyed by all the hh.o drama
May 26 17:35:55 <goxboxlive> zecke: So there isent any problems to run the oz354x on the htc universal? I thougt that branch was made for the Zaurus.
May 26 17:35:58 <emte> why? was there more this morn?
May 26 17:36:32 <hrw> re..
May 26 17:37:14 <hrw> fscking desktop hang
May 26 17:38:58 <hrw> hang during patching kernel.. so remove, unpack another time ;(
May 26 17:43:16 <danboid> goxboxlive: linux kernel doesn't even support charging or suspend/resume on the uni last I heard
May 26 17:43:45 <hrw> danboid: so?
May 26 17:44:25 <goxboxlive> danboid: I know, it's charging, but there are fun to just play around with it. 
May 26 17:44:31 <hrw> danboid: linux kernel does not even support charging on collie
May 26 17:44:59 <emte> se things get developed by people who work on supporting the device
May 26 17:45:31 <goxboxlive> danboid: We have made an wiki at handhelds. Here you can see our progress: http://www.handhelds.org/moin/moin.cgi/Universal
May 26 17:49:30 <zecke> kergoth: ho, not only you
May 26 17:50:20 <zecke> goxboxlive: you are free to leverage from the work of the OpenZaurus people (hrw!)
May 26 17:50:51 <hrw> ;)
May 26 17:51:46 <koen> zecke: stop using 'leverage' as a verb
May 26 17:52:57 <zecke> koen: well, I can use it as a verb (when making it transitive)
May 26 17:53:42 <goxboxlive> well as you allready have seen so is my english so bad that i even dident understood the meaning of it :-)
May 26 18:07:37 <CosmicPenguin> NOTE: package livecd-image-1.0: completed
May 26 18:07:38 <CosmicPenguin> yay!
May 26 18:07:47 <koen> cool
May 26 18:08:15 <CosmicPenguin> I still need to figure out a smart way to do initrds with OE
May 26 18:08:30 <zecke> CosmicPenguin: ask RP, he is doing that for qemuarm
May 26 18:08:39 <danboid> whats this, an oe live CD for x86?
May 26 18:08:47 <CosmicPenguin> yeah - for Geode
May 26 18:08:56 <CosmicPenguin> Well, actually its the infrastructure
May 26 18:09:08 <CosmicPenguin> I'll leave the actual content to more customer oriented types
May 26 18:12:08 <hrw> ~lart linux kernel devs
May 26 18:12:08 * ibot blasts linux kernel devs with a huge firehose then strangles linux kernel devs with it
May 26 18:12:14 <hrw> arch/i386/kernel/signal.c: In function ?do_signal?:
May 26 18:12:14 <hrw> arch/i386/kernel/signal.c:587: error: label ?no_signal? used but not defined
May 26 18:12:23 <hrw> 2.6.17-rc5
May 26 18:15:25 <emte>  2235 emte      15   0 1852m 1.3g  848 S  1.0 67.4 158:19.24 xmms
May 26 18:15:25 <emte>        you think xmms might have a memory leak?
May 26 18:16:06 * danboid look, hrw, xmms- strike! :)
May 26 18:16:17 <hrw> danboid: what is xmms?
May 26 18:16:57 <danboid> ?
May 26 18:17:10 <emte> my system started to misbehave ~ the same time hrw came back from his lockup
May 26 18:17:29 <emte> you passing bad juju around today hrw?
May 26 18:17:37 <danboid> your chance to slag off xmms- not interested?
May 26 18:17:54 <hrw> emte: yes
May 26 18:18:14 <hrw> emte: http://www.hrw.one.pl/2006/05/26/amd64-switch-to-it-or-not/ is a reason probably
May 26 18:18:31 * emte reads
May 26 18:21:06 <emte> hrw, why not just cluster your other two until after the wedding?
May 26 18:21:23 <emte> not that two is really a cluster ...
May 26 18:21:29 <hrw> emte: 1700+ is at work, 22+ is at home
May 26 18:21:50 <emte> clustering does not require the machines in the same geographical area
May 26 18:22:07 <emte> but your work might not like it ...
May 26 18:22:13 <hrw> home machine is too unstable ;(
May 26 18:23:11 * koen should try to usurp some more uni machines
May 26 18:26:41 <goxboxlive> I need some help again: When i have edited the machine.conf file ( for my machine htcuniversal.conf ) how do i make sure taht the new bootstsrap-image has my changes in it. There isent enough with just bootstrap-image -cclean and the same for task-bootstrap. Should i delete some files in the work directory?
May 26 18:27:25 <woglinde> goxbox if you really want sure delete the tmp-directory
May 26 18:28:50 <goxboxlive> But id i delete the tmp directory doesnt it have to build all again? Like glibc, QT, fakerrot etc? That takes so much time. Isent there another solution to it?
May 26 18:29:18 <hrw> goxboxlive: rebuild task-bootstrap
May 26 18:30:02 <goxboxlive> hrw: I guess i have to rebuild the xanadux-un-2.6.bb file also then. THX
May 26 18:30:52 <hrw> probably not
May 26 18:31:17 <goxboxlive> ok
May 26 18:31:49 <goxboxlive> hrw: Ok, i was to fast, i rebuilt it, but it's building so fast anyway.
May 26 18:35:20 <cbrake> how does one install gdb-cross on the host machine?  I've built it w/ bitbake, but can't find any evidence that it is installed anywhere.
May 26 18:36:13 <hrw> check tmp/staging/
May 26 18:36:22 <koen> or cross/ ?
May 26 18:38:45 <hrw> it goes into /usr/local/$SDK_NAME
May 26 18:39:44 <cbrake> not in cross or staging ...
May 26 18:40:07 <hrw> tmp/deploy/sdk/
May 26 18:40:20 <cbrake> hrw: ahh
May 26 18:41:02 <cbrake> hrw: nope, not in there either, do I need to install the ipkg or something ??
May 26 18:41:16 <cbrake> hrw: or is there some option for sdk generation?
May 26 18:41:59 <hrw> hmm.. have to check probably
May 26 18:47:32 <koen> zecke: could you make a correct patch for the glibc -isystem troubles?
May 26 18:48:10 <zecke> 'correct'?
May 26 18:48:20 <zecke> koen: workaround as in using -I for glibc
May 26 18:48:25 <koen> yes
May 26 18:48:33 <zecke> koen: or correct as in switching glibc to schilling make?
May 26 18:48:45 <goxboxlive> Another problem: My bootstrap-image does only create 'ttyS0' I also need ttyS1 (bluetooth, bluez-utils is made) and ttyS2 (irda, irattach is made). What to do?
May 26 18:48:55 <koen> zecke: if you want :)
May 26 18:50:01 <hrw> goxboxlive: 2.6 kernel? it should be done by udev
May 26 18:50:59 <zecke>  koen: hmm someone masquerading on OSX is broken :}
May 26 18:51:26 <goxboxlive> hrw: Yes, 2.6 kernel. Ok, i'll define udev. Thanks alot.
May 26 18:53:23 * cbrake manually copies arm-linux-gdb from ipkg for now until he figures out how sdks work ...
May 26 18:57:07 <hrw> cbrake: Packaged contents of gdb-cross into /home/hrw/devel/build/test/tmp/deploy/ipk/gdb-cross_6.3-r0_x86_64.ipk
May 26 19:01:05 <cbrake> hrw: nod, looks about the same as the one I built.
May 26 19:01:42 <hrw> cbrake: you build for mipsel? :)
May 26 19:04:00 <cbrake> hrw: it seems that bitbaking gdb should also drop a copy in tmp/cross/bin for convenience.
May 26 19:04:14 <cbrake> hrw: no, still building for ARM, but everything else looks the same :-)
May 26 19:04:37 <cbrake> hrw: what type of device is the mips build for?
May 26 19:04:48 <hrw> wrt54
May 26 19:05:50 <cbrake> hrw: nice -- something else to try .... someday :-)
May 26 19:06:37 <hrw> cbrake: OE does not build working distro for wrt54 but can build packages for openwrt ;)
May 26 19:07:21 <cbrake> hrw: IC
May 26 19:07:42 <cbrake> hrw: BTW, I got a patch failure when building cherokee 0.5.3 this morning is this something you would expect?
May 26 19:08:04 <hrw> will recheck
May 26 19:08:07 <kergoth> adding the wrt kernel was to be a pain in the ass. lots of references to code outside the kernel tree which were of questionable licensing..
May 26 19:08:37 <Crofton> ~lart bubious licensing
May 26 19:08:38 * ibot gives bubious licensing a "free" copy of Windows and then charges double for "Upgrades"
May 26 19:08:45 <cbrake> hrw: http://pastebin.com/739902
May 26 19:09:15 <zecke> One license to rule them all
May 26 19:09:30 <Crofton> in the land of Redmond
May 26 19:11:30 <emte> zecke, isnt that MS's definition of "it's" invention of Open Source?
May 26 19:12:29 <zecke> emte: well they have ~6 shared source licenses
May 26 19:13:30 <zecke> emte: I thought more of GPLv2 or BSD
May 26 19:14:15 <emte> well MS claims to have invented the idea/concept
May 26 19:14:22 <emte> like everything else
May 26 19:14:43 <emte> and if they can trademark or patenet it they will
May 26 19:14:49 <emte> patent*
May 26 19:16:59 <reenoo> evening
May 26 19:17:08 <zecke> hey
May 26 19:18:38 <hrw> cbrake: hmm.. I recheckout cherokee on dev. and it work
May 26 19:19:29 <Ken|JLime> I want to map sh3 against sh, when compiling pcmcia-cs. Any idea on how to do this?
May 26 19:19:50 <Ken|JLime> in other words make it know that sh3 is an sh arch
May 26 19:19:54 <woglinde> hi renoo
May 26 19:20:11 <cbrake> hrw: I'll check my build tree again ...
May 26 19:20:13 <hrw> Ken|JLime: TARGET_ARCH_sh3 = "sh"?
May 26 19:20:14 <reenoo> hey woglinde 
May 26 19:20:32 <Ken|JLime> hrw: doh
May 26 19:23:40 <Ken|JLime> hrw: creates another set of trouble, I'll just make a patch for the configure script
May 26 19:41:50 <CosmicPenguin> http://hg.fedoraproject.org/hg/olpc/tools/image--devel?f=66af776534b4;file=conf/packages-core.txt
May 26 19:53:00 <CosmicPenguin> what the heck is sugar anyway?
May 26 19:53:14 <zecke> CosmicPenguin: their gui
May 26 19:53:20 <zecke> CosmicPenguin: based on XUL runner
May 26 19:53:55 <CosmicPenguin> goo
May 26 19:55:01 <zecke> CosmicPenguin: http://www.0xdeadbeef.com/weblog/?p=200
May 26 20:00:47 <zecke> koen: did you add monotone 0.26?
May 26 20:03:40 <CoreDump|home> omg, there are still nice human beeing out there. My neighbour (who I don't know _at_ _all_) just handed my back my swiss army knife (~40?)I lost at my parking lot
May 26 20:04:13 <zecke> CoreDump|home: that can happen
May 26 20:04:50 <CoreDump|home> how high is the chance of that these days ;)
May 26 20:05:07 <mickeyl> low
May 26 20:05:17 <CoreDump|home> ~lart WinXP samba handling
May 26 20:05:17 * ibot breaks out the Hoover and sucks up WinXP samba handling
May 26 20:05:18 <woglinde> corredump depends were you live
May 26 20:05:19 <zecke> at the parking lot?
May 26 20:05:21 <zecke> high
May 26 20:05:30 <zecke> only my cat could steal it from there...
May 26 20:05:38 <CoreDump|home> zecke: =)
May 26 20:05:42 <woglinde> coredump in other cities you where stabbed by your oen knife
May 26 20:05:54 <CoreDump|home> woglinde: lol, sad but true
May 26 20:06:02 <zecke> woglinde: or killed because you have a mp3 player
May 26 20:06:13 <zecke> So get a ogg player!
May 26 20:06:14 <mickeyl> :(
May 26 20:06:22 <zecke> an ogg...
May 26 20:12:10 <lkaine> Is the SL-5500 supported much anymore?
May 26 20:12:17 <lkaine> Collie that is
May 26 20:12:32 <hrw> is
May 26 20:13:30 <CoreDump|home> heh sure is
May 26 20:14:13 <Zero_Chaos> lkaine: yeah, it is a nice device
May 26 20:16:08 <cbrake> hrw: cherokee working now.  I must have had some stale data in the cache due to a package in a separate bb collections tree I had deleted.
May 26 20:18:02 <hrw> cbrake: bb collections.. have to try it one day
May 26 20:18:06 <hrw> collie suxx
May 26 20:18:27 <CosmicPenguin> bbcollections are so nice
May 26 20:18:40 * hrw will exchange collie for h22xx
May 26 20:19:10 <hrw> I usually use two entried in BBFILES
May 26 20:19:18 <lkaine> the ipaq eh?
May 26 20:19:30 <hrw> lkaine: yep
May 26 20:19:46 <lkaine> why so
May 26 20:20:21 <hrw> lkaine: because collie is a crap device
May 26 20:20:44 <thejapa> marcin: hi! but h2200 has no integrated network
May 26 20:20:57 <hrw> thejapa: so do collie
May 26 20:21:10 <lkaine> hrw:  Well after coming back to OZ after a couple years absence and having it sit in a box doing nothing...I'm trying to put it to good use
May 26 20:21:26 <hrw> thejapa: nice to see you Marcos here
May 26 20:21:50 <hrw> lkaine: I use my collie to collect dust and testing oz before release
May 26 20:21:53 <thejapa> :) i tried like hell to make a sandisk wifi work in my ipaq, only to plug it into a notebook and find it is broken
May 26 20:22:09 <thejapa> if I could go for another ipaq, i'd try one with wifi :)
May 26 20:22:22 <thejapa> h2200 has bluetooth, better than nothing
May 26 20:22:23 <lkaine> is the CL-3200 good?
May 26 20:22:42 <hrw> lkaine: any cxx00 is nice
May 26 20:22:49 <ixoth> my dream = SL-5600 + integrated BT + wifi
May 26 20:22:52 <hrw> atleast I heard that from users
May 26 20:23:08 <hrw> ixoth: get sl-6000 and add bt
May 26 20:23:16 <thejapa> marcin: i'm trying to find rpm for monotone 0.25 for fedora 5 right now 
May 26 20:23:26 <lkaine> hrw: how much are cxx00?
May 26 20:23:29 <thejapa> i'll end up installing debian on my notebook
May 26 20:23:33 <ixoth> it's larger and has lower baterry life
May 26 20:23:54 <hrw> lkaine: c1000 (no microdrive) is 300-350 eur
May 26 20:24:11 <hrw> thejapa: rebuilt it from source
May 26 20:24:35 <lkaine> hrw: aren't the c series with built-in hd much better so that you don't have to worry about space problems?
May 26 20:24:49 <thejapa> ok
May 26 20:25:09 <hrw> lkaine: I boot my c760 from 512M sd card
May 26 20:26:16 <lkaine> can you run kde3.5 lite with openzaurus?  I was reading a thread about people that were successful using it with the cxx00 series on pdaXrom
May 26 20:26:19 <ixoth> oh my... built-in hd is for installing gentoo probably... the idea of HD in PDA is stupid, IMHO
May 26 20:26:33 <ixoth> oh my.... KDE on Zaurii...
May 26 20:26:41 * ixoth faints
May 26 20:27:21 <lkaine> yeah that's what I read
May 26 20:27:24 <lkaine> it looked really nice
May 26 20:27:35 <hrw> kde on 64M device with only vga screen... its madness not usage
May 26 20:28:07 <hrw> one day I was forced to use 800x600 monitor - after 0.5h with kde I switched to console
May 26 20:29:02 <JoeSchmo> fluxbox would be a nice addition to openzaurus, i've been usiing it for a while, it's prety nice
May 26 20:29:16 <lkaine> I had a similar problem.  I got the supertux package working after ipkg screwed the links and packages so I had to iterate through every library making sure it was linked correctly and then I finally got it working without sound...and then the screen on my collie wasn't big enough to display it
May 26 20:29:42 <CoreDump|home> lkaine: =)
May 26 20:29:45 <lkaine> right now I have my collie running x11/matchbox
May 26 20:30:11 <ixoth> I hope I'll never have to move from Opie
May 26 20:30:12 * CoreDump|home prefers Opie on QVGA
May 26 20:30:28 <zecke> hehe
May 26 20:30:42 <zecke> I hope to have Qtopia4 on my Things soon(tm)
May 26 20:30:56 <JoeSchmo> I'd use opie if snes9x or zsnes worked through it
May 26 20:30:57 <lkaine> I like GPE better
May 26 20:31:00 <hrw> zecke: I'm waiting for opie/x11
May 26 20:31:09 <zecke> hrw: make menuconfig ;)
May 26 20:31:22 <zecke> hrw: matchbox got support for Ok/Cancel in the decoration
May 26 20:31:35 <hrw> zecke: qt2/x11 is ugly...
May 26 20:31:47 <hrw> zecke: when opie will get it back?
May 26 20:31:48 <lkaine> hrw: i agree
May 26 20:32:04 <zecke> hrw: what back?
May 26 20:32:16 <hrw> zecke: ok/cansel in decoration
May 26 20:33:45 <zecke> hrw: sure, Opie hints to show Ok/Cancel
May 26 20:34:03 <lkaine> a couple years ago I helped develop OZ...but now everything is different :P
May 26 20:34:17 <zecke> hrw: matchbox informs Opie ok or cancel were pressed
May 26 20:34:47 <ixoth> I've even wrote app for the first Z contest, hehe... now the old days are gone...
May 26 20:35:08 <hrw> ixoth: there are users of such old apps still
May 26 20:35:31 * chouimat is away: pub
May 26 20:36:26 <lkaine> I helped fix the kernel for a SD card io problem...and an issue with hot reboot
May 26 20:36:35 <lkaine> can't remember exactly though which...been awhile
May 26 20:36:49 <ixoth> my app is in OE, btw... shame for me i never managed to get it work good :)
May 26 20:36:58 <lpotter_> ixoth: which app?
May 26 20:37:21 <ixoth> qlot
May 26 20:37:29 <ixoth> qplot, that's it
May 26 20:37:51 <CIA-9> 03mickeyl 07org.oe.dev * rbe0f8737... 10/conf/machine/include/motorola-ezx.conf: update
May 26 20:38:41 <hrw> hmm.. when last time I build oz kernel on this machine....
May 26 20:39:07 * hrw want amd64....
May 26 20:40:04 <woglinde> hrw buy one
May 26 20:40:09 * lkaine has amd64....
May 26 20:40:23 * emte doesnt want amd64
May 26 20:40:35 * ixoth wants mips64
May 26 20:40:43 * woglinde wants hal9000
May 26 20:40:58 <lkaine> I've got a 4 processor P4 Xeon machine with 2.5gb ECC w/2 20gb 15k SCSIs
May 26 20:41:11 * zecke wants AMDX2 on AM2 with 4800+
May 26 20:41:18 <hrw> woglinde: sooner or later will do - first have to collect some cash to open own business, travel to few places, marry Ania and then amd64 maybe
May 26 20:41:23 <hrw> zecke: and 4GB ram?
May 26 20:41:24 <lkaine> zecke: yeah it's nice...my roommate has one
May 26 20:41:28 * cbrake likes amd64 a lot :-)
May 26 20:41:33 * woglinde really wants a iseries 590
May 26 20:41:48 <lkaine> what I want to see is a low voltage Duo-core processor on a PDA...
May 26 20:41:55 <hrw> cvI like your amd64 too ;)
May 26 20:42:00 <emte> altho in reality ... if your going to buy a new board you may want to wait ~ 4-6mo
May 26 20:42:05 <hrw> lkaine: get omap one ;)
May 26 20:42:05 <zecke> hrw: 4?
May 26 20:42:05 <zecke> hrw: 4GB as 1st level cache
May 26 20:42:29 <emte> the new 8Gb memory chipsets should be availible
May 26 20:42:29 <hrw> zecke: then why amd64 x2? get 4xopt875
May 26 20:43:03 <CoreDump|home> the x2 is a tad little bit cheaper ;)
May 26 20:43:28 <lkaine> right now i'm running a P-M 2.26ghz laptop w/2gb DDR2 100gb hd...Nvidia 7800 GTX video card 17" 1920x1200 native res trubrite UXGA :P
May 26 20:43:56 <emte> sounds like a damn heavy backache
May 26 20:44:07 <lkaine> not really...it's rather light for me  
May 26 20:44:10 <woglinde> lkain with 30 min for akkulifetime?
May 26 20:44:33 <hrw> I use nearly 4 years old crt1 17" 1280x960
May 26 20:44:36 <ixoth> and I am running Sharp MM20 laptop 1Ghz 512RAM 20Gb - it's 950g
May 26 20:44:41 <lkaine> I have 2 1/2 hours while playing a DVD... >3 hours when writing documents or doing anything else not using the video card
May 26 20:45:57 <woglinde> hrw I have a over 10 years old ibm 19"
May 26 20:46:05 <woglinde> with 1024x1024
May 26 20:46:11 <emte> they made 19" 10 yrs ago?
May 26 20:46:18 <lkaine> wow
May 26 20:46:34 <woglinde> ibm 6091 19i
May 26 20:46:40 <zecke> woglinde: and I get a headache sitting in front of it
May 26 20:46:47 <emte> mmm 96 ... yeah they started increasing monitor sizes around 94 ish ...
May 26 20:46:48 <woglinde> why?
May 26 20:46:49 <zecke> woglinde: and I swear this monitor eats little kitten
May 26 20:46:51 <woglinde> 75 hz
May 26 20:46:54 <likewise> Sony Multiscan 400PS 19" and not going away before it breaks
May 26 20:47:06 <hrw> woglinde: I see flicker in 75Hz
May 26 20:47:11 <woglinde> yes sony manufactred it for ibm
May 26 20:47:25 <emte> hrw, you have flourecent lights in the room?
May 26 20:47:27 <lkaine> that's amazing...
May 26 20:47:28 <woglinde> hrw get glasses
May 26 20:48:01 <hrw> woglinde I wear glasses for over 10 years
May 26 20:48:07 <hrw> sorry for msg
May 26 20:48:08 <woglinde> hrw :)
May 26 20:48:11 <hrw> ircing remote
May 26 20:48:39 <emte> flourecent lights flicker in the 70-80 Hz range and you notice it when your monitor is close to the same frequency
May 26 20:48:51 <lkaine> whenever I run opie...it seems a lot slower than GPE counterpart.  I am running a 64mb memory with no storage on flash zImage configuration
May 26 20:48:55 <hrw> few years glasses (stronger and stronger) then contacts then glasses back
May 26 20:49:42 <zecke> lkaine: all operations? or just start feels slower?
May 26 20:49:45 <emte> i always have to change the monitors at school to 65Hz or 85Hz, i cant stand any range between em. too big a headache
May 26 20:50:10 <lkaine> zecke:  mostly all operations..and especially the package manager...it's REALLY slow
May 26 20:50:11 <woglinde> emte yes there are many monitor's az 60 hz
May 26 20:50:22 <woglinde> problem windows default settings
May 26 20:50:32 <woglinde> and the people dont know how to change it
May 26 20:50:35 <emte> no all the systems are cloned
May 26 20:50:40 <woglinde> or that it can be changed
May 26 20:50:40 <lkaine> yeah..my lcd here runs @ 60hz
May 26 20:50:47 <woglinde> lkain bah
May 26 20:50:52 <lkaine> ?
May 26 20:50:52 <woglinde> lcd no prob
May 26 20:51:00 <woglinde> crt with 60hz
May 26 20:51:01 <emte> and whoever makes the initial image probably doesnt care
May 26 20:51:03 <lkaine> ah..yea
May 26 20:51:49 <woglinde> have to change it at my mothers work
May 26 20:52:10 <lkaine> which distro has more packages vs. performance..opie or gpe?
May 26 20:52:18 <emte> opie
May 26 20:52:34 <emte> packages made for it *
May 26 20:52:35 <zecke> lkaine: well the package manager is bad to compare
May 26 20:53:03 <emte> gpe has the benifits of being able to use most x11 apps
May 26 20:53:22 <ixoth> but the overhead of X...
May 26 20:53:42 <ixoth> and most X11 apps aren't suitable to use on 320x240
May 26 20:53:56 <ixoth> or even on 640x480
May 26 20:53:57 <emte> most are usuable with a few modifications
May 26 20:54:03 <TheCan> ixoth, well you have to be picky about this
May 26 20:54:04 <lkaine> i have a collie...so 320x240
May 26 20:54:24 <TheCan> <- has a jornada, so 640x240
May 26 20:54:46 <emte> being able to export your desktop as well has it's benifits
May 26 20:54:47 <hrw> ~fsck agere.com
May 26 20:55:04 <ibot> e2fsck /dev/agere.com : warning! filesystem contains helpdesk workers!
May 26 20:55:17 <lkaine> I noticed a problem with gpe/x11 is that the fonts are much huger..I just want something I can use to possibly surf wifi/read pdf documents/etc.  Would opie be best for this?
May 26 20:55:42 <woglinde> opie is okay for this
May 26 20:55:56 <woglinde> with the new konq-emb
May 26 20:56:31 <emte> konq-emb builds for gpe as well does it not?
May 26 20:56:41 <woglinde> emte yes
May 26 20:56:46 <woglinde> sorry
May 26 20:57:35 <lkaine> is qpdf available for opie?  
May 26 20:58:22 <ixoth> and btw, what's the browser for new Qtopia? opera again?
May 26 20:58:49 <emte> THATS IT!! i have found programmers worse than MS
May 26 20:58:58 <emte> they are microchip!
May 26 20:59:29 <lkaine> microchip?
May 26 20:59:30 <emte> they all need to be sent back to school to learn how to manage code
May 26 21:00:10 <lkaine> does openzaurus use Ant/Maven or any type of continuous integration?
May 26 21:00:24 <emte> if you want an example of the worst programming parcitces on earth, look at the Microchip enc28j60 tcpstack/webserver
May 26 21:00:41 <emte> practices*
May 26 21:01:20 <hrw> emte: friend implemented vfat on pic18 with 1.5K ram
May 26 21:01:29 <lkaine> I have to deal with bad programming practices every day...unfortunately I have to correct them.  At least I get paid for it.
May 26 21:01:41 <emte> i hope the vfat code is cleaner than this
May 26 21:01:52 <hrw> emte: had to be
May 26 21:02:21 <woglinde> hrw did he paided ms for patent?
May 26 21:03:18 <emte> this one has structures and unions strewen throught ~30 files there are full functions in headers and function declarations in c files, and everything inter-includes everything else because nothing is in the right place
May 26 21:04:37 <emte> there are function pointers to function pointers rhat point to functions bound with unions to data typecast returns
May 26 21:04:38 <lkaine> emte: yuck
May 26 21:04:49 <lkaine> I repeat previous statement
May 26 21:05:38 <thejapa> newbie question -> what version of boost should be used with monotone? my fc5 has boost-devel.i386 1.33.1-5, but monotone 0.25 is complaining Error: Missing Dependency: libboost_regex.so.1 is needed by package monotone
May 26 21:05:39 <thejapa> Error: Missing Dependency: libboost_filesystem.so.1 is needed by package monotone
May 26 21:05:39 <thejapa> Error: Missing Dependency: libboost_date_time.so.1 is needed by package monotone
May 26 21:06:01 <thejapa> marcin: the sourceforge is same version, guess i need an older one?
May 26 21:06:02 <hrw> Version: 0.25-0.1
May 26 21:06:03 <hrw> Depends: libboost-date-time1.33.1, libboost-filesystem1.33.1, libboost-regex1.33.1, libc6 (>= 2.3.5-1), libgcc1 (>= 1:4.1.0), libstdc++6 (>= 4.0.2-4), zlib1g (>= 1:1.2.1)
May 26 21:06:15 <hrw> thejapa: thats from my debian box
May 26 21:06:22 <emte> thejapa, did you install the boost extensions or just the core library?
May 26 21:06:34 <zecke> thejapa: try the static monotone binary
May 26 21:06:47 <thejapa> both, i guess. on fc5 it's boost.i386 and boost-devel.
May 26 21:06:55 <ixoth> mine;s[  ] dev-libs/boost-1.33.1 (1)
May 26 21:07:04 <emte> you need bost-filesystem boost-date  etc
May 26 21:07:15 <emte> bost-filesystem*
May 26 21:07:21 <thejapa> emte: thanks
May 26 21:07:22 <emte> bah i still cant spell it
May 26 21:07:32 <thejapa> was about to throw fedora out the windows
May 26 21:07:34 <thejapa> oops
May 26 21:08:03 <lkaine> is there an easy way to move your rootfs to an sd card?  I want the flexibility of not having to worry about "no space left on device"
May 26 21:08:15 <zecke> lkaine: altboot from CoreDump|home 
May 26 21:08:30 <CosmicPenguin> They have SD cards with infinite space now?  leet
May 26 21:08:47 <lkaine> haha
May 26 21:08:55 <woglinde> cosmic 2 gig orr 4 gig shoud neer inifinite
May 26 21:09:01 <zecke> CosmicPenguin: using tape
May 26 21:09:01 <emte> CosmicPenguin, if you never plan to read any data back, yeah :)
May 26 21:09:33 <lkaine> CosmicPenguin: the problem I was having was that I would do a 64mb install...and then I would install a few packages and then my rootfs would get filled up with symbolic links or whatever and ipkg would...not obey the -d sd and put everything there
May 26 21:09:47 <zecke>  koen: do you use Apple's X-Server?
May 26 21:09:57 <emte> my infinity drive is convieniently mapped to /dev/void 
May 26 21:10:28 <lkaine> and my drive that will eventually collapse space-time is also mapped to /dev/null
May 26 21:10:54 <emte> hmm
May 26 21:11:15 <emte> speaking of that ... i wonder which stage the accelerator is at
May 26 21:12:14 <thejapa> can u dd if=/dev/void of=/dev/null?
May 26 21:12:15 <thejapa> duh
May 26 21:14:05 <hrw> fsck.. hermes-II wifi cards are supported only under 2.4 ;(
May 26 21:15:30 <emte> hmm cant find the page ...
May 26 21:15:43 <emte> oh well the earth implodes next year anyway
May 26 21:15:47 <thejapa>  rpm -ql boost.i386 | egrep "(regex|filesystem|date_time)"
May 26 21:15:48 <thejapa> /usr/lib/libboost_date_time.so.1.33.1
May 26 21:15:48 <thejapa> /usr/lib/libboost_date_time.so.2
May 26 21:15:48 <thejapa> /usr/lib/libboost_filesystem.so.1.33.1
May 26 21:15:48 <thejapa> /usr/lib/libboost_filesystem.so.2
May 26 21:15:48 <thejapa> /usr/lib/libboost_regex.so.1.33.1
May 26 21:15:50 <thejapa> /usr/lib/libboost_regex.so.2
May 26 21:15:59 <thejapa> can I softlink them??
May 26 21:16:30 <woglinde> emte no the meteroit should collapse yesterday into earth
May 26 21:17:44 <emte> woglinde, no, the final stage of the LHC CERN Accelerator goes active then, one of the thoeries is that earth will collapse
May 26 21:18:05 <emte> ofcourse they say that about every new accelerator
May 26 21:31:23 <thejapa> ok
May 26 21:31:26 <thejapa> didn't work
May 26 21:31:54 <thejapa> i'm gonna use debian
May 26 21:33:44 <zecke> thejapa: using the static built?
May 26 21:34:17 <zecke> wb back master
May 26 21:34:43 <pb_> good evening zecke
May 26 21:34:45 <hrw> ~curse agere
May 26 21:34:47 <hrw> hi pb
May 26 21:34:50 <ibot> May you be reincarnated as a Windows XP administrator, agere !
May 26 21:35:01 <pb_> hi hrw
May 26 21:35:11 <woglinde> hi pb
May 26 21:35:22 <woglinde> pb did the pub close?
May 26 21:35:30 <hrw> 802.11g cards which use Agere Hermes II chipset are supported only under 2.4 ;(
May 26 21:35:54 <woglinde> hrw no chance to port the driver?
May 26 21:36:04 <woglinde> to the new wifi-architecture
May 26 21:36:11 <thejapa> zecke: no, standard rpm 
May 26 21:36:26 <hrw> woglinde: source is available on GPL so someone can start it
May 26 21:36:29 <thejapa> er... how do I get it static? sorry for the lame question
May 26 21:36:36 <pb_> hi woglinde 
May 26 21:36:39 <zecke> thejapa: it is static
May 26 21:36:49 <zecke> thejapa: put it in your home dir ;)
May 26 21:36:49 <pb_> actually no, the pubs are still open
May 26 21:36:56 <woglinde> thejappa compile monotone your self
May 26 21:37:07 <woglinde> pb :)
May 26 21:37:11 <thejapa> yeah, master told me that already :)
May 26 21:37:28 <thejapa> i was giving it a last try... 
May 26 21:37:30 <hrw> thejapa: http://venge.net/monotone/downloads/monotone-0.25.2-linux-x86.bz2
May 26 21:37:33 <woglinde> than go to work *yedimovement*
May 26 21:38:04 <zecke> woglinde: the young padawan...
May 26 21:38:14 <zecke> woglinde: get me some finish food *yedimovement*
May 26 21:38:20 <zecke> woglinde: and some nachos *yedimovement*
May 26 21:38:32 <pb_> heh
May 26 21:38:40 <thejapa> :D
May 26 21:38:55 <zecke> hmm where is that finish coming from...
May 26 21:38:58 * zecke feels manipulated
May 26 21:40:01 <pb_> I'll be on vacation for a few days from tomorrow.  back on wednesday.
May 26 21:40:13 * pb_ taking a short trip to the frozen north
May 26 21:40:24 <hrw> pb_: enjoy
May 26 21:40:30 <Ken|JLime> anyone built lzo successfully lately?
May 26 21:40:45 <hrw> Ken|JLime: I did - for arm on amd64 iirc
May 26 21:41:07 <thejapa> ppl, thanks for the help, i must go to my operating systems class :0
May 26 21:41:14 * hrw got few oopses here so no new app can be started..
May 26 21:41:32 <hrw> thejapa: cu then
May 26 21:41:38 <thejapa> :)
May 26 21:41:47 <thejapa> cu in pt-br is assho*e
May 26 21:41:52 <woglinde> pb scotland?
May 26 21:41:57 <thejapa> i mean, the "organ"
May 26 21:41:59 <thejapa> or hole
May 26 21:42:00 <thejapa> :)
May 26 21:42:04 <thejapa> bye
May 26 21:42:08 <hrw> thejapa: good to know
May 26 21:42:12 <woglinde> bye hrw
May 26 21:42:14 <pb_> woglinde: not quite that far north, but nearly
May 26 21:42:19 <Ken|JLime> hrw: oki, means I must bugtrack abit further then
May 26 21:42:22 <woglinde> loch ness
May 26 21:42:24 <woglinde> *g*
May 26 21:42:42 <pb_> newcastle, northumberland, that kind of area
May 26 21:42:49 <hrw> woglinde: the question is... to see a monster or to be a monster... D:
May 26 21:43:32 <woglinde> okay will go to sleep too
May 26 21:43:54 <pb_> woglinde: night night
May 26 21:43:55 <hrw> yep.. time to reboot, start few torrents and go sleep
May 26 21:44:25 <hrw> brb
May 26 21:48:46 <hrw> re
May 26 21:52:33 <lkaine> does openzaurus use cvs or svn?
May 26 21:53:05 <hrw> monotone
May 26 21:58:40 <mickeyl> looks we're slowly getting somewhere...
May 26 21:58:41 <mickeyl> http://www.vanille.de/temp/a780/DSC02830.JPG
May 26 21:58:42 <mickeyl> g'night
May 26 21:59:15 <zecke> mickey|zzZZzz: hmm we should consider buying a Qtopia phone license :}
May 26 21:59:26 <zecke> but nice!
May 26 22:01:21 <reenoo> mickey|zzZZzz: cute! congrats!
May 26 22:01:56 <hrw> mickey|zzZZzz: nice progress!
May 26 22:02:55 <ixoth> wow, nice!
May 26 22:10:24 <Ken|JLime> mickey|zzZZzz: gratz :D
May 26 22:10:36 <lkaine> what pda is that?
May 26 22:10:45 <Ken|JLime> :-P
May 26 22:10:51 <hrw> lkaine: its phone - motorola a780
May 26 22:11:01 <ixoth> that's motorola ezx phone
May 26 22:12:01 <zecke> koen: the world graph is down to a 3.1mb png
May 26 22:12:24 <lkaine> i see thanks
May 26 22:15:49 <lkaine> did he cross compile oz for the a780, or through the openezx project?
May 26 22:16:28 <zecke> lkaine: neither
May 26 22:16:38 <lkaine> zecke: then what?
May 26 22:16:49 <zecke> lkaine: He uses OpenEmbedded to create an image/distribution for the A780
May 26 22:16:58 <zecke> OZ!=OE
May 26 22:17:07 <lkaine> ok i see
May 26 22:17:25 <lkaine> OE maintains specific kernel drives for different architectures or what?
May 26 22:17:28 <lkaine> *drivers
May 26 22:17:46 <zecke> lkaine: no, OE is that what Debian doesn't want to be
May 26 22:17:57 <zecke> lkaine: we are a toolkit, specialised on Embedded and mobile linux
May 26 22:18:04 <zecke> to cross compile and create distributions
May 26 22:18:20 <zecke> OpenZaurus is using OE to create their distribution
May 26 22:18:21 <lkaine> ah.  when you say what Debian doesn't want to be, what do you mean?
May 26 22:18:30 <lkaine> ah ok
May 26 22:18:43 <zecke> lkaine: see the recent reports from the DebConf
May 26 22:18:58 <lkaine> a specific distribution aimed for the Zaurus (collie, c series, 5600 etc)?
May 26 22:18:58 <zecke> lkaine: OpenEmbedded is just a toolkit to create stuff
May 26 22:20:02 <lkaine> it maintains the libraries and such for creating stuff?  i'm sorry.  I'm trying to get into this and haven't been into it in awhile.
May 26 22:20:07 <ixoth> OpenEmbedded is like ports system in BSD
May 26 22:20:13 <ixoth> or portage in Gentoo
May 26 22:20:24 <lkaine> I use Debian/FreeBSD..so I know
May 26 22:20:36 <zecke> not  like that
May 26 22:20:56 <ixoth> hmmm... similar... we have metadata and compile it
May 26 22:20:57 <zecke> lkaine: OpenEmbedded is a set of metadata (recipes) to build and create stuff (e.g. flashable images, feeds)
May 26 22:21:03 <lkaine> but what does OE actually in terms of creating images etc?
May 26 22:21:09 <lkaine> ok
May 26 22:21:27 <zecke> lkaine: OpenEmbedded is so fleixble projects like OpenZaurus, Familiar or Angstrom
May 26 22:21:37 <zecke> can use it to create their distributions
May 26 22:21:43 <lkaine> and OZ creates the kernel for the specific architecture (collie, 5600 etc) and then uses OE to create the zImage and initrd.bin?
May 26 22:21:50 <zecke> lkaine: 'creates'
May 26 22:22:02 <lkaine> well, maintains/extends
May 26 22:22:03 <zecke> lkaine: RP and do13 write the kernels
May 26 22:22:28 <zecke> lkaine: the border is 'thin' as most of the OZ devels (hrw!) are OE devels as well
May 26 22:22:36 <zecke> so we 'share' the recipes...
May 26 22:22:42 <ixoth> for example, I use OE to make images for x86 and MIPS routers
May 26 22:23:04 <hrw> and recent OZ fix for samba was made by people working on SH4 machines
May 26 22:23:08 <lkaine> does OpenWRT also use OE?
May 26 22:23:23 <zecke> lkaine: not that I know
May 26 22:23:28 <hrw> lkaine: no - but Oe can be used to build packages for openwrt
May 26 22:23:52 <hrw> I use openwrt on my wrt54gs and sometimes I install OE built stuff there
May 26 22:24:07 <lkaine> hrw: I use openwrt as well on my wrt54g
May 26 22:24:32 <Ken|JLime> I use it on my jornada 680 :D (SuperH-3 Based)
May 26 22:24:39 <lkaine> zecke: when you say RP and do13 write the kernels, these are specific developers on the oz project?
May 26 22:25:14 <hrw> lkaine: they are OE devels
May 26 22:25:16 <hrw> rather then OZ devels
May 26 22:25:45 <hrw> most of devels working in OZ project work on OE
May 26 22:25:52 <lkaine> so OZ maintains specific packages or kernel modules relevant to OZ specific stuff only?
May 26 22:26:09 <zecke> lkaine: OZ selects specific versions of packages
May 26 22:26:17 <zecke> lkaine: to build their distribution
May 26 22:26:23 <zecke> lkaine: e.g. a version of the gcc
May 26 22:26:28 <Ken|JLime> lkaine: its like gentoo, it grabs the sources, patches it and builds it
May 26 22:26:48 <zecke> but is more flexible than like Gentoo, Ports, buildroot
May 26 22:26:52 <Ken|JLime> yeah
May 26 22:26:52 <hrw> lkaine: OZ people make some packages changed to work them a bit different with OZ (like checkin kernel version on each boot)
May 26 22:27:46 <lkaine> so what project actually handles the specific kernel writing that's specific to zaurus hardware only, such as the collie, 5600, cx00, etc?
May 26 22:28:02 <lkaine> or is that OE still?
May 26 22:28:06 <hrw> lkaine: OE with help of OZ devels
May 26 22:28:37 <hrw> lkaine: Z running linux-openzaurus 2.6 from OE can boot openzaurus, familiar, angstrom, umbaumba distros
May 26 22:30:03 <lkaine> hrw: ok, I think I am understanding.  
May 26 22:31:03 <hrw> lkaine: OE contain metadata, OZ use metadata to build distro which can be run on any machine which is in OE and is arm
May 26 22:31:49 <lkaine> hrw: OE only supports StrongARM arch?
May 26 22:32:00 <lkaine> hrw: or xscale?
May 26 22:32:10 <ixoth> OE supports everything :)
May 26 22:32:21 <ixoth> doesn't really care about arches
May 26 22:32:36 <lkaine> but OZ only builds for arm/xscale or what?
May 26 22:32:46 <ixoth> you can add description to generate packages for your own machine
May 26 22:32:57 <hrw> oe support: x86 (few flavours), mipsel, powerpc, arm (xscale, strongarm, arm920, arm926)
May 26 22:33:14 <Ken|JLime> *chuckle*
May 26 22:33:36 <hrw> OZ is built for strongarm (feed) and xscale (most of machines)
May 26 22:33:42 <zecke> lkaine: in contrast to other solutions OE can generate anything, any package format, any arch
May 26 22:33:59 <Ken|JLime> hrw: dont forget the SuuuuuperH arch also :D
May 26 22:34:02 <hrw> angstrom will be built for xscale, iwmmxt, crunch models of arm
May 26 22:34:07 <hrw> Ken|JLime: sorry..
May 26 22:34:11 <Ken|JLime> hrw: :P
May 26 22:34:15 <lkaine> and pdaxrom uses OE?
May 26 22:34:17 <hrw> oe also support SuperH: 3 and 4
May 26 22:34:21 <hrw> lkaine: no
May 26 22:34:24 <Ken|JLime> :)
May 26 22:34:31 <lkaine> what about pocketdebian?
May 26 22:34:46 <Ken|JLime> embedded debian?
May 26 22:34:46 <zecke> no
May 26 22:34:47 <hrw> we also support sparcs
May 26 22:36:49 <lkaine> so if I wanted to create/build/maintain a package in the feed...I would setup an OE build environment..and then setup OZ build environment etc?
May 26 22:37:14 <zecke> lkaine: you would get OE
May 26 22:37:23 <zecke> lkaine: configure it using the OpenZaurus settings
May 26 22:37:26 <zecke> lkaine: write a bb file
May 26 22:37:31 <zecke> lkaine: and build it
May 26 22:37:48 <zecke> lkaine: http://oe.handhelds.org/screencasts/org.openembedded.htm
May 26 22:39:28 <lkaine> okay, watching the video
May 26 22:41:54 <hrw> Ken|JLime: lzo built for arm
May 26 22:42:42 <Ken|JLime> hrw: I fixed it by setting ac_cv_sizeof_char_p=....=4, but still have no idea why it built for arm and x86
May 26 22:43:52 <Ken|JLime> Got http://bugs.openembedded.org/show_bug.cgi?id=1050, if you want to do a quick look
May 26 22:43:57 <lkaine> what I want to know is why isn't OE using Ant or Maven?
May 26 22:44:16 <zecke> lkaine: ant? Java build tool?
May 26 22:44:16 <hrw> ant is that 'make for java'?
May 26 22:45:23 <hrw> java makes me sick
May 26 22:45:27 <lkaine> Ant is a build management tool.  It's far superior than make and provides configuration through verbose metadata and build rules specified through XML.  Maven uses Ant to provide further functionality through plugins and allow building easier.
May 26 22:45:39 <hrw>  2163 hrw       15   0  423m 148m  27m S  8.7 16.8   5:10.01 java
May 26 22:45:51 <lkaine> Ant/Maven are written in java, but are not solely for java projeccts.
May 26 22:46:04 <Ken|JLime> Sounds like a good way to get new bugs
May 26 22:46:06 <zecke> lkaine: well, can it do OVERRIDES easily?
May 26 22:46:13 <hrw> anything which is java written has -999 points here
May 26 22:46:19 <koen> zecke: yes I added monotone 0.26 to OE
May 26 22:46:20 <zecke> e.g. without writing 1000 lines of XML to apply a patch
May 26 22:46:28 <koen> or at least I had a working .bb in my tree
May 26 22:46:34 <zecke> only for machine FOO and distri MOO?
May 26 22:46:40 <zecke> koen: okay, you miss a patch?!
May 26 22:46:49 <lkaine> zecke: it can handle OVERRIDES easily
May 26 22:46:53 <koen> zecke: I did?
May 26 22:47:02 <zecke> koen: looks like, not buildable here
May 26 22:47:16 <zecke> lkaine: can it do collections?
May 26 22:47:23 <hrw> Ken|JLime: Packaged contents of lzo into /home/hrw/devel/build/test/tmp/deploy/ipk/lzo_1.08-r14_mipsel.ipk
May 26 22:47:33 <lkaine> zecke: check it out if you want.  I'm sure it does.
May 26 22:47:44 <zecke> lkaine: lol
May 26 22:48:01 <Ken|JLime> hrw: :(
May 26 22:48:18 <Ken|JLime> hrw: How can it build for all those without needing an ac_cv_sizeof_char_p??
May 26 22:48:46 <zecke> lkaine: so you have not used ant to realize common OpenEmbedded use cases and wonder why it is not used?
May 26 22:48:47 <hrw> lkaine: I think that we should switch to ptxdist - it even has 'make menuconfig' so you do not have write config file to start build
May 26 22:49:01 <hrw> Ken|JLime: dont ask me...
May 26 22:49:04 <zecke> hrw: I always wanted to propose that
May 26 22:49:41 <hrw> zecke: ptxdist or mae menuconfig?
May 26 22:49:58 <zecke> hrw: ptxdist
May 26 22:50:21 <lkaine> does everyone hate java here or what?  lol
May 26 22:50:29 <Ken|JLime> probobly
May 26 22:50:43 <hrw> lkaine: I hate everything which take 0.5G ram for nothing
May 26 22:50:48 <zecke> lkaine: no? just answer the questions
May 26 22:51:10 <zecke> lkaine: so you have not realized the OE uses cases in ant, why do you think it is comparable to OE?
May 26 22:51:23 <koen> aaaargh
May 26 22:51:24 <koen> ant
May 26 22:51:53 <koen> I still have a ant.bbclass under a very icky license on ewi
May 26 22:51:53 <zecke> lkaine: OE are thousands of tasks (in our language BBfiles/recipes)
May 26 22:51:56 <hrw> let someone port EAutoBuild from AmigaOs... or EMake...
May 26 22:52:20 * CoreDump|home chuckles
May 26 22:52:33 <zecke> http://ewi546.ewi.utwente.nl/OE/opie-build.png
May 26 22:52:43 <hrw> I used EAutoBuild to generate copies for each user of my MultiView
May 26 22:52:44 <lkaine> it was just a suggestion.  I asked earlier if OE/OZ had any type of continuous integration system in place.
May 26 22:53:03 <Ken|JLime> zecke: haha
May 26 22:53:10 <zecke> 'continous' integration?
May 26 22:53:32 <hrw> zecke: nice graph
May 26 22:53:32 <zecke> lkaine: if you suggest FOO technology, you should know FOO really well and what it will replace ;)
May 26 22:53:34 <lkaine> zecke: you've never heard the term before?
May 26 22:53:50 <zecke> lkaine: well my CS language is german
May 26 22:54:33 <lkaine> zecke: what?
May 26 22:54:33 <koen> zecke: cute
May 26 22:54:46 <ggilbert_> lkaine: His primary language is german.
May 26 22:55:04 <zecke> http://ewi546.ewi.utwente.nl/OE/gpe-build.png
May 26 22:55:17 <zecke> My Computer Science terminology is german
May 26 22:55:31 <zecke> lkaine: so if you get me a proper translation, I'm sure I know what you mean
May 26 22:55:43 <hrw> Continuous Integration is a software development practice where members of a team integrate their work frequently, usually each person integrates at least daily - leading to multiple integrations per day. Each integration is verified by an automated build (including test) to detect integration errors as quickly as possible. Many teams find that this approach leads to significantly reduced integration problems and allows a team to develop cohesive software 
May 26 22:55:50 <hrw> http://www.martinfowler.com/articles/continuousIntegration.html
May 26 22:56:08 <ggilbert_> what's the german word for spam? :p
May 26 22:56:13 <lkaine> I was almost finished typing, but yeah...
May 26 22:56:27 <reenoo> ggilbert_: spam
May 26 22:56:32 <koen> zecke: I think we can flag pkgconfig-native as a bogus dep as well 
May 26 22:56:32 <hrw> ggilbert_: Polish word is 'mielonka' (meat spam)
May 26 22:56:51 <ggilbert_> hehe
May 26 22:56:54 <Ken|JLime> ggilbert_: swedish word is 'spam'
May 26 22:57:04 <ggilbert_> argh. It was a joke folks :p
May 26 22:57:16 <Ken|JLime> :-P
May 26 22:57:19 <lkaine> so I take it OE/OZ doesn't use cont. integration.  
May 26 22:57:28 <ggilbert_> In hawaii they just call it delicious
May 26 22:57:38 <Ken|JLime> ggilbert_: hehe
May 26 22:58:09 <zecke> lkaine: we have a tinderbox, but it is not as easy as one expects
May 26 22:58:15 <hrw> lkaine: if'Continuous Integration is a software development practice where members of a team integrate their work frequently' is your definition then we do it
May 26 22:58:38 <lkaine> hrw: no, that's not it
May 26 22:58:41 * CoreDump|home looks at the GPE graph and thinks: Bloody hell
May 26 22:58:45 <hrw> lkaine: but we do not have too much autobuilds
May 26 22:58:51 <zecke> koen: check the graph again
May 26 22:59:07 <lkaine> hrw: it's not so much as integrating work frequently, but building automatically 24/7 to provide the facility of providing less bugs and more communication between developers of a project
May 26 22:59:35 <zecke> lkaine: tinderbox
May 26 22:59:47 <koen> zecke: much better
May 26 22:59:55 <reenoo> continuous integration relies on automated tests. OE doesn't cover that type of thing for the most part
May 26 22:59:59 <zecke> lkaine: but the issue is how to test ~10.000 interactive (GUI) and non-interative tools
May 26 23:00:30 <zecke> lkaine: but I guess ant is magically solving this case
May 26 23:00:44 <lkaine> I think this is a dead end...people here are too closed minded. 
May 26 23:00:45 <zecke> so yeah pxdist and ant looks perfetcly right
May 26 23:00:59 <zecke> lol
May 26 23:01:24 <ixoth> good luck everyone...
May 26 23:01:30 <ixoth> time for me to sleep
May 26 23:01:32 <zecke> lkaine: I just think you have no domain knowledge
May 26 23:01:47 <hrw> lkaine: basically OE could be automated to do tests - iirc nslu people do autobuilds for their targets
May 26 23:02:00 <reenoo> heh
May 26 23:02:11 <reenoo> autobuilds != strict unit tests
May 26 23:02:23 <lkaine> yep
May 26 23:02:39 <hrw> anyway  its 01:00 here so I rather have most of brain suspended sobetter will shut
May 26 23:03:03 <zecke> koen: now added gpe-calendar and datebook to the graph
May 26 23:03:27 <zecke> why is gtk+ needing libgcrypt to build?
May 26 23:04:07 <koen> zecke: display migration?
May 26 23:04:27 <zecke> koen: gtk+ doesn't link against display migration?!
May 26 23:04:36 <zecke> koen: and displaymigration is even secure?
May 26 23:05:00 <koen> the gcrypt dep has been under discussion a while ago
May 26 23:05:29 <zecke> http://ewi546.ewi.utwente.nl/OE/matchbox-build.png
May 26 23:05:56 <lkaine> zecke: and I do have domain knowledge...I was only curious as to what standards were in place...and continuous integration isn't specific domain knowledge..it's basic se knowledge.  unit and mock tests could provide a facility that could be useful.  and ant doesn't magically solve it, you have to do the work yourself
May 26 23:05:59 <zecke> matchbox-panel depends on virtual/kernel?
May 26 23:06:26 <koen> zecke: wireless tools plugins?
May 26 23:06:36 <koen> anyway
May 26 23:06:39 <hrw> koen: they should use libiw rather?
May 26 23:06:40 * koen -> sleep
May 26 23:06:42 <zecke> lkaine: well, I doubt you have domain knowledge on toolchains, cross compiling, packaging
May 26 23:07:16 <zecke> lkaine: and well, we start to get Continous building and static analysis in place (slowly)
May 26 23:07:19 <lkaine> zecke: not much, no.  I am trying to learn and was asking questions.  
May 26 23:07:33 <zecke> lkaine: the next part will involve dynamic analysis (runtime testing)
May 26 23:07:47 <lkaine> wtf mate
May 26 23:08:01 <lkaine> zecke: what is your problem?
May 26 23:08:07 <zecke> none
May 26 23:08:17 <zecke> I'm telling you what is there, what is getting build
May 26 23:08:21 <Ken|JLime> koen: hell just frooze over, Cause I built opie-image for the first time :-D
May 26 23:08:44 <ggilbert_> lkaine: I suspect your going to run into manpower problems with any testing beyond verifying that stuff builds.
May 26 23:09:11 <zecke> Ken|JLime: wow, Opie conquers SuperH
May 26 23:10:21 <ggilbert_> lkaine: Zecke runs a build daemon to make sure a few targets build
May 26 23:10:25 <hrw> Jornada ODevice?
May 26 23:10:44 <ggilbert_> but in terms of complete testing, that just doesn't happen because of resource limiations.
May 26 23:10:46 <lkaine> ggilbert_: testing doesn't need involve actual testing.  through strict test-driven development, unit testing, etc...the tests would be automated and wouldn't need involving anyone except for writing tests.  
May 26 23:11:07 <zecke> Builds, Source is available, patches apply and common errors
May 26 23:11:08 <Ken|JLime> zecke: yeah :D
May 26 23:11:08 <lkaine> and when I say testing.  I don't mean actual testing.  I mean unit tests or mock tests.
May 26 23:11:21 <ggilbert_> lkaine: Well tests would need to be written for several thousand bb's would they not?
May 26 23:11:21 <hrw> cu
May 26 23:11:43 <ggilbert_> and as I said, a major roadblock would just be the ability to build everything fequently
May 26 23:11:54 <lkaine> that's true
May 26 23:12:33 <ggilbert_> I haven't built this stuff in a while, but a single target was around 6 hours or so
May 26 23:12:50 <zecke> lkaine: http://liw.iki.fi/liw/talks/finnish-inquisition.pdf
May 26 23:13:14 <ggilbert_> maybe more actually, kind of fuzzy
May 26 23:13:16 <zecke> lkaine: What we can do today is: Compile regulary in many configurations for many targets
May 26 23:13:46 <ggilbert_> zecke: how long does a full world build take these days?
May 26 23:14:01 <zecke> check if source-code is available, check if patches are present and apply, check if the recipes match common quality criteria
May 26 23:14:17 <zecke> when compiling packages we do basic quality checking already or soon will do
May 26 23:14:34 <zecke> what we currently don't do: Testing the packages on the target automatically
May 26 23:14:58 <zecke> lkaine: as this would mean having unit-tests for KDE, GNOME, Linux Kernel
May 26 23:15:01 <zecke> ...
May 26 23:15:16 <zecke> ggilbert_: days probably
May 26 23:15:18 <zecke> ggilbert_: weeks ;)
May 26 23:15:59 <ggilbert_> well you don't need to test every package in depth
May 26 23:16:44 <zecke> lkaine: we had bugs like that: zip was supposed to be compiled for arm
May 26 23:16:50 <lkaine> and you don't need to build continuously on every package, only the central repository
May 26 23:17:16 <zecke> lkaine: but was compiled using the host compiler, using 'file' on package building can uncover these issues
May 26 23:17:23 <zecke> lkaine: well, in OE there is no center
May 26 23:17:35 <lkaine> monotone is distributed right?
May 26 23:17:50 <lkaine> or rather OE is distributed versioning via monotone
May 26 23:18:03 <zecke> lkaine: in theory, yes
May 26 23:18:03 <ggilbert_> lkaine: not sure what you're getting at
May 26 23:18:48 <ggilbert_> Building the "blessed" tree? That's still an extremely large build with quite a few combinations
May 26 23:19:22 <zecke> lkaine: monotone allows to attach the results of test cases to the recipes
May 26 23:19:51 <zecke> lkaine: so in the future we will attach the results of automatic anaylsis automatically
May 26 23:21:17 <Ken|JLime> zecke: just booted it up, works wonderfully :D (although in need of some adjusting for the j6xx)
May 26 23:21:22 <lkaine> zecke: for future reference I wasn't saying it should switch.  I was only asking and questioning.  
May 26 23:21:23 <zecke> Ken|JLime: wow
May 26 23:21:30 <zecke> lkaine: okay
May 26 23:21:42 <zecke> lkaine: but as with everything where humans are involved
May 26 23:21:45 <ggilbert_> lkaine: Not sure if we're disconnected somewhere, but I'm talking about a single  build of whatever you want to consider the official tree
May 26 23:21:49 <zecke> it is one thing what information you can get
May 26 23:21:54 <zecke> and what information will be used
May 26 23:22:07 <zecke> lkaine: http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=OpenEmbeddedBuilds
May 26 23:24:05 <lkaine> zecke: this is the amount of time for building a specific tree/image?
May 26 23:24:36 <zecke> lkaine: on a ~fast p4 a build of a GUI image (Opie or GPE) takes roughly four hours
May 26 23:24:52 <zecke> lkaine: we can do clobbered builds (rebuilding the stuff that changed)
May 26 23:25:14 <zecke> lkaine: and what I'm working on ATM is creating a Bonsai for Monotone and SVN
May 26 23:25:34 <zecke> lkaine: you can leave comments on the build reports etc...
May 26 23:25:36 <ggilbert_> jebus, what are the nslu2 guys building on?
May 26 23:26:01 <zecke> ggilbert_: that is a BitBake bug :}
May 26 23:26:10 <zecke> ggilbert_: it misses to inform the box it is done
May 26 23:26:18 <ggilbert_> hehe
May 26 23:26:22 <lkaine> I was about to say...20 days to do 1 build.
May 26 23:26:51 <ggilbert_> I've always been under the opinion that a custom buildd might be handy for  oe
May 26 23:27:12 <ggilbert_> tinderbox just doesn't seem to get to a very interesting level of granularity
May 26 23:27:29 <zecke> ggilbert_: yet! ;)
May 26 23:27:33 <ggilbert_> hehe
May 26 23:27:40 <zecke> as everything it depends on the available time
May 26 23:27:44 <ggilbert_> yeah
May 26 23:27:53 <ggilbert_> Which is why I never finished writing that buildd
May 26 23:28:11 <zecke> Vision: It will provide a query page. You will be able to do queries like
May 26 23:28:27 <zecke> Show me all versions of FOO that got build on platforms xyz using gcc xyz
May 26 23:28:46 <zecke> and each build report will get a sane overview as well
May 26 23:28:57 <zecke> http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=29&logfile=20060526005538.log
May 26 23:29:00 <lkaine> zecke: what are you writing it in?  perl?
May 26 23:29:04 <ggilbert_> I'd like to see  what packages didn't build :p
May 26 23:29:09 <zecke> lkaine: sadly it is perl :(
May 26 23:29:26 <zecke> ggilbert_: 'build' = were attempted to build
May 26 23:29:27 <lkaine> zecke: perl isn't too bad.  I know perl myself...but prefer php...and j2ee.
May 26 23:29:46 <zecke> I'm a python/c++ guy
May 26 23:29:51 <zecke> doing Java only at Uni
May 26 23:29:52 <ggilbert_> python's a bit handy here. Easier to integrate with bitbake.
May 26 23:30:10 <zecke> I like the Java stuff the Apache team created...
May 26 23:30:30 <zecke> the object bridge would be perfectly suited for the tinderbox
May 26 23:30:31 <lkaine> zecke: Jakarta?
May 26 23:31:12 <zecke> lkaine: not only that. But their mapping tool (Object Bridge or such)
May 26 23:31:12 <zecke> lkaine: you write you data model, run the tool and specify the database
May 26 23:31:12 <zecke> lkaine: and it will manage the database communication for you
May 26 23:31:28 <lkaine> zecke: you mean hibernate/JDBC?
May 26 23:31:38 <zecke> no :)
May 26 23:31:47 <zecke> hibernate is not from Apache?!
May 26 23:31:54 <ggilbert_> fucking metacity
May 26 23:32:00 <lkaine> zecke: I'm not sure what you're referring to
May 26 23:32:19 <lkaine> zecke: iBATIS SQL maps maybe?
May 26 23:33:12 <lkaine> zecke:  oh...is it HiveMind?
May 26 23:33:14 <zecke> lkaine: Struts+Tomcat+ http://db.apache.org/ojb/
May 26 23:33:24 <lkaine> zecke: Struts...yuck
May 26 23:33:25 <zecke> makes a perfect server application
May 26 23:33:31 <lkaine> no it doesnt :P
May 26 23:33:48 <zecke> ojb is wonderful though
May 26 23:33:50 <lkaine> Struts is ugly.  even more metadata to deal with..and the implementation isn't as good.  Rumor is that Struts is going dead.
May 26 23:34:06 <zecke> I might be mistaken about struts then
May 26 23:34:13 <lkaine> obj is nice..but I'd rather use hibernate
May 26 23:34:17 <zecke> but ojb was nice to use
May 26 23:34:30 <lkaine> if you want to look into a better framework check out http://www.springframework.org/
May 26 23:34:41 <zecke> but yeah, the tinderbox is written and perl
May 26 23:34:52 <zecke> and it is too much to decide to do better in another language
May 26 23:35:01 <lkaine> too late you mean?
May 26 23:35:28 <zecke> hehe, well in proper english :)
May 26 23:36:00 <zecke> too much as in == Too much code to simply throw away and start something from scratch
May 26 23:36:08 <lkaine> zecke: well, your english is better than my german...I speak almost none :P lol
May 26 23:36:14 <lkaine> zecke: I understand
May 26 23:37:01 <zecke> lkaine: the 'time' didn't weight too much. Using a badly written perl application allows us to use it 'now'
May 26 23:37:21 <zecke> a proper Java Application which is well designed would take way longer
May 26 23:37:31 <lkaine> zecke: I understand.  PHP is the same way.
May 26 23:39:30 <zecke> obviously you are welcome to help on automatic testing and establishing a useful framework
May 26 23:43:10 <lkaine> are you saying the perl app or what?
May 26 23:43:52 <zecke> anything 
May 26 23:44:16 <reenoo> 'night all
May 26 23:44:47 <zecke> lkaine: are you a student?
May 26 23:44:49 <lkaine> a lot of programmers in europe everyone is going to bed.. :P
May 26 23:44:50 <zecke> reenoo: sleep well!
May 26 23:44:58 <lkaine> zecke: always
May 26 23:45:03 <zecke> I'm going to bed soon as well
May 26 23:45:14 <zecke> lkaine: are you enrolled in a university?
May 26 23:45:45 <lkaine> zecke: yes.  going for my bachelors in computer science and mathematics
May 26 23:46:20 <lkaine> zecke: you?
May 26 23:47:03 <zecke> lkaine: CS and Physics (Diplom somewhere inbetween master and bachelor)
May 26 23:47:46 <lkaine> zecke: I'll be a student for awhile yet...I'm going for my Ph.D. eventually
May 26 23:48:50 <zecke> 'awhile' probably applies to me as well
May 27 00:02:01 <zecke> koen: configure.ac-no-sync-with-stdio-0.26.patch is missing
May 27 00:02:41 <kergoth> This D language is nifty.  should add its bits to oe and use it to build the dev env, since 90% of tools dont support it out of the box yet
May 27 00:03:05 <zecke> koen: I'm not too excited yet ;)
May 27 00:03:18 <zecke> kergoth: I'm not too excited yet
May 27 00:03:26 <zecke> kergoth: you can mix C and the successor C++
May 27 00:03:38 <kergoth> and you can mix C and D as well
May 27 00:03:43 <kergoth> just not C++ and D
May 27 00:03:44 <zecke> kergothi: and you can mix C++ and the successor D? No :}
May 27 00:03:57 <zecke> that is bad, then it should be called objective C 
May 27 00:04:04 <zecke> oh that already exists :}
May 27 00:04:12 <kergoth> which is as it should be.  they want to ditch the backwards compatibility nightmare
May 27 00:04:23 <kergoth> not to mention there's no standard c++ ABI
May 27 00:06:01 <kergoth> as i understand it, one of the big issues is that its quite difficult to parse c+++ without implementing 90% of a c++ compiler
May 27 00:06:04 * kergoth shrugs
May 27 00:06:13 <zecke> kergoth: hehe
May 27 00:06:38 <zecke> kergoth: yeah, compared to Java c++ is horrible
May 27 00:06:46 <kergoth> someone wrote a java to D translator
May 27 00:06:51 <kergoth> apparently its much easier than c++
May 27 00:06:53 <kergoth> hehe
May 27 00:07:06 <JoeSchmo> depends if you can memorize long names or not
May 27 00:07:10 <zecke> kergoth: there is a reason gcc implemented a hand written c++ parser
May 27 00:07:14 <kergoth> hehe
May 27 00:08:43 <JoeSchmo> system.out.go.to.hell.console.print.string "Hello World!";
May 27 00:09:57 <kergoth> D is quite nice though, it gives you a lot of the tools you need to rapidly develop while retaining a C degree of power and performance.  unfortunately teh language isnt stable, and as you say, lacking c++ compatibility might be a barrier to entry
May 27 00:10:38 * kergoth really likes that it ditched c++ multiple inheritence (a pile of crap imo) in favor of both interfaces and mixins
May 27 00:10:56 <zecke> kergoth: mixins?
May 27 00:11:21 <kergoth> see the wikipedia definition.  its the most common usage of multiple inheritence that isnt just interfaces
May 27 00:11:23 <lkaine> multiple inheritance = nightmare
May 27 00:11:46 <kergoth> "mixing in" common functionality into classes
May 27 00:12:39 <emte> hmm
May 27 00:12:44 <zecke> hmm too tired :)
May 27 00:12:44 <emte> STL has it uses
May 27 00:12:44 <kergoth> define template Foo containing methods a and b, do a 'mixin Foo' in the class definition, and now that class has those methods
May 27 00:14:00 <emte> its the only place i can think of offhand where i use multiple inheritance ... 
May 27 00:14:41 <emte> but i havent looked at D myself
May 27 00:15:06 <kergoth> check out the example code on http://d-programming-language.org/ to see the syntax and some of the features
May 27 00:15:19 <emte> too buzy trying to plot how to send Microchip's programmers back to school
May 27 00:15:40 <zecke> kergoth: I find a common virtual machine and ABI more interesting
May 27 00:15:44 <lkaine> D seems a bit like C#...hrmm....
May 27 00:15:48 <zecke> compile once
May 27 00:15:55 <zecke> optimize everywhere
May 27 00:16:16 <emte> thats machinecode tho
May 27 00:16:36 <zecke> emte: sure, just another stack machine
May 27 00:17:36 <emte> you can optimize for the machine, the code itself, or ideally both
May 27 00:17:38 <lkaine> seems interesting. 
May 27 00:17:58 <emte> peopel usually settle for the first two
May 27 00:18:07 <emte> since that generates more money
May 27 00:19:33 <emte> ugh  argspecs().count
May 27 00:19:35 <zecke> I have not explored JITs yet
May 27 00:19:48 <emte> i hate it when people write that way
May 27 00:21:30 <zecke> anyway good nite
May 27 00:21:45 <emte> altho it seems easier to follow it in D
May 27 00:22:46 <emte> probably because argspecs() doesnt repointed to multiple times
May 27 00:23:41 <kergoth> the 'auto' keyword is kinda cool.  basically does a heap allocation for the object, not the stack, but it gets automatically freed at the end of that scope, rather than being gc'd
May 27 00:23:44 <lkaine> Instead of D they should have called it "napl"  for "not another programming language"
May 27 00:25:11 <Ken|JLime> calling it a nite, cya tommorow
May 27 00:27:19 <emte> lkaine, i belive that already exists
May 27 00:28:36 <emte> The programming language for. SYNAPSE1 is called nAPL neural Algorithms Pro-. gramming Language
May 27 00:29:13 <lkaine> ok fine...
May 27 00:29:21 <lkaine> it was a joke anyway.. :P
May 27 00:30:27 <kergoth> zecke should read http://www.digitalmars.com/d/faq.html#call-cpp-code .. hehe
May 27 02:04:13 <likewise> Djeez... Up at 3:45 AM... 
May 27 02:05:04 <CIA-9> 03lenehan 07org.oe.dev * r58268aaf... 10/packages/net-snmp/net-snmp_5.1.2.bb: 
May 27 02:05:04 <CIA-9> net-snmp 5.1.2: Fix endianess detection to work on all platforms by looking
May 27 02:05:04 <CIA-9> at the site file to determine the correct configure parameter (inspired by
May 27 02:05:04 <CIA-9> how openssl does this). Previously this built on mipsel only. Should build
May 27 02:05:04 <CIA-9> on all platforms now. Closes bug #483.
May 27 02:05:08 <CIA-9> 03lenehan 07org.oe.dev * r36ffefda... 10/packages/net-snmp/net-snmp_5.1.2.bb: net-snmp 5.1.2: Does not work with a parallel build, so clear PARALLEL_MAKE. Closes bug #713.
May 27 02:18:15 <CIA-9> 03lenehan 07org.oe.dev * r7a03dd20... 10/packages/hdparm/hdparm_6.3.bb: hdparm 6.3: Update SRC_URI to point at the hdparm primary site which is now sourceforge. Closes bug #772.
**** ENDING LOGGING AT Sat May 27 02:59:56 2006