**** BEGIN LOGGING AT Sun Jun 12 23:59:57 2005 Jun 13 00:00:36 <_law_> has anyone any idea? http://pastebin.ca/14233 Jun 13 01:36:54 morning Jun 13 01:37:35 morning Jun 13 01:38:07 good morning all Jun 13 01:42:19 hi proti & koen Jun 13 01:42:53 Zero_Chaos: see bug #47 for a work-around Jun 13 01:46:00 Has anyone ever booted root from NFS? Is it any useful? I'm thinking about adding it to altboot but only if it makes sense. Jun 13 01:46:19 It is very useful Jun 13 01:46:25 morning, btw. Jun 13 01:46:30 morning mickeyl Jun 13 01:46:31 hey mickeyl Jun 13 01:46:54 CoreDump|home: almost all 2.6 ports for for ipaqs use nfsroot over usbnet Jun 13 01:47:15 koen: ah i see Jun 13 01:47:16 03mickeyl 07 * r1.3564 10openembedded/packages/qte/ (qte_2.3.10.bb qte-2.3.10/c7x0-w100-accel.patch): qte-2.3.10: add c7x0 w100 accelleration bits Jun 13 01:47:19 03mickeyl 07 * r1.3563 10openembedded/packages/sharp-binary-only/sharp-aticore-oss_1.0.1.bb: ship aticore-oss-examples in a dedicated package Jun 13 01:47:20 so I concur with mickeyl: it is very useful Jun 13 01:47:47 * CoreDump|home starts hacking Jun 13 01:47:50 mickeyl: w100 accell is 'fluff' Jun 13 01:48:26 koen: :D Jun 13 01:49:55 oi Jun 13 01:50:33 If anyone wants to play with it, set QWS_DISPLAY=W100:0 Jun 13 01:50:45 cool. Jun 13 01:50:52 hi all Jun 13 01:50:54 whats the license on that patch? Jun 13 01:51:04 djeez, the code got bloated by 6kb Jun 13 01:51:12 and it isn't pim Jun 13 01:51:17 dunno, need to ask SirFred on that Jun 13 01:51:18 mickeyl: in /etc/init.d/opie script? Jun 13 01:51:27 hrw|work: ya Jun 13 01:52:21 RP,mithro: ping Jun 13 01:53:57 ljp: aticore is result of reassembler so who know... Jun 13 01:55:19 you have to differenciate between aticore and the patch to support aticore in qte Jun 13 01:55:35 using dlopen it may be licensed as BSD, but i'm not sure Jun 13 01:56:39 aticore itself is highly problematic Jun 13 01:56:56 heh, so is ATI Jun 13 01:57:04 ya Jun 13 01:57:38 not only would they not give us infos, but they wouldnt let SHarp give us infos either Jun 13 01:57:50 nice Jun 13 02:03:44 hmm, that qte patch didnt show up ;( Jun 13 02:05:08 http://www.linuxdevices.com/news/NS4665550218.html Jun 13 02:07:20 ya, thats works. i help test the betas a bit Jun 13 02:16:17 weird.. showed up on my other box Jun 13 02:16:24 03koen 07 * r1.3562.1.2 10openembedded/conf/machine/h3600.conf: h3600.conf: guard jffs image args with a _jffs2 Jun 13 02:19:46 hmm.. tomorrow will be good day to build newer OE stuff Jun 13 02:22:02 someone sent me the knights sources Jun 13 02:22:28 ljp: bk pull - we mirror SRC_URI on openzaurus.org Jun 13 02:23:28 well, the sources werent even on oesources Jun 13 02:23:40 the distros should have them Jun 13 02:23:56 anyone who distributes the knight binary should have them Jun 13 02:28:07 ljp: I have them. I'll upload the archive, hold on Jun 13 02:28:21 morning all Jun 13 02:28:33 hi RP Jun 13 02:28:51 mickey|busy: pong Jun 13 02:29:52 ljp: http://hentges.net/tmp/knights.tar.gz Jun 13 02:30:29 aitlib will hopefully be able to go away sometime soon. I just need time... Jun 13 02:30:43 morning RP Jun 13 02:47:15 re Jun 13 02:50:02 hi pb Jun 13 02:50:12 ~lart Debian kernel-package Jun 13 02:50:12 * ibot strangles Debian kernel-package with a doohicky mouse cord Jun 13 02:50:19 /usr/src/linux/scripts/gcc-version.sh: line 12: i486-linux-gnu-gcc: command not found Jun 13 02:50:29 hi hrw|work Jun 13 02:51:53 someone in debian decided that now there will be i486-linux-gnu-gcc but gcc-3.3 (debian default) is not yet updated.. Jun 13 02:52:12 argh Jun 13 02:56:10 pb_, initial indications are that the glibc-fp-byteorder.patch works with glibc-2.3.5 Jun 13 02:56:23 jacques: rock Jun 13 02:56:30 pb_, just replace all /sysdeps/ with /ports/sysdeps/ in the patch Jun 13 02:56:36 righto Jun 13 02:56:47 I'll go directly to my CVS client. Jun 13 02:56:51 :-D Jun 13 03:00:14 pb_: wasn't the tslib patch correct? it's still not committed :( Jun 13 03:02:38 I didn't have time to look at it yet. If you could submit a new patch, with the redundant bits (i.e. the parts that just duplicate h3600 files) removed, and everything in one patch rather than a patch and a tarball, that'd help. Jun 13 03:07:11 w00t, first NFS boot complete :) Jun 13 03:07:38 CoreDump|home: congrats Jun 13 03:07:45 thx Jun 13 03:09:18 mhcln01:/[...]/OpenEmbedded/tmp/rootfs on / type nfs (rw,v2,rsize=8192,wsize=8192,hard,udp,lock,addr=mhcln01) Jun 13 03:09:39 i thought it would be much slower Jun 13 03:09:54 CoreDump|home: nfs over usbnet? wifi? Jun 13 03:09:59 wifi Jun 13 03:10:27 I have to find time and install openwrt on router. linksys firmware suxx to badly Jun 13 03:10:31 need to dig out my craddle to implement the USB stuff Jun 13 03:11:24 first I have to get 100% working wifi setup to forget about usb cable Jun 13 03:11:26 lol, even suspend / resume works Jun 13 03:11:53 jacques: right, the patch is in Jun 13 03:11:55 sort of Jun 13 03:12:23 CoreDump|home: what transfer you have over wifi? Jun 13 03:12:50 hrw|work: about 300-350KB/sec IIRC Jun 13 03:13:33 heh.. I really have to get wrt working better Jun 13 03:13:42 50-60KB/s is all I got Jun 13 03:13:47 hmm odd, the hardware keys are dead after a suspend/resume cycle. Jun 13 03:13:48 pb_, thanks! Jun 13 03:13:52 hrw|work: ouch Jun 13 03:14:16 and i though 300KB sucks Jun 13 03:14:47 CoreDump|home: so whe I want to move some stuff to Z I prefer card reader, then usbnet, then wifi (when works) then bt Jun 13 03:15:36 makes sense. Did you try switching to another channel? Jun 13 03:17:30 CoreDump|home: all of them Jun 13 03:17:34 doh Jun 13 03:17:47 11 worked last week, 13 worked Jun 13 03:17:55 maybe your WiFi card is b0rked? Jun 13 03:18:10 did you try channel 6? Jun 13 03:18:16 yesterday none worked so I want to flash something which can be managed remotely via shell Jun 13 03:18:36 ok - I have to go home turn my pc.. be back in ~30m Jun 13 03:21:00 pb_, are you glibc arm maintainer now/again ? Jun 13 03:35:57 jacques: yeah Jun 13 03:36:47 pb_, fantastic :-) Jun 13 03:38:30 pb_: did you tell RMK the sky has stopped falling down? Jun 13 03:39:18 pb_: done. I say it's ready to commit now: http://bugs.openembedded.org/show_bug.cgi?id=66 Jun 13 03:43:32 ibot: botmail for zecke: StevieH kann heute Nachmittag nicht. Jun 13 03:47:58 re Jun 13 03:48:29 hrw|work: wb Jun 13 03:50:55 pb_: That's good news. Jun 13 03:53:30 RP: what's that? Jun 13 03:57:54 koen: heh, no. Jun 13 03:58:01 Arjan: very good. Jun 13 03:59:15 ~curse amarok Jun 13 03:59:17 May you be reincarnated as a Windows XP administrator, amarok ! Jun 13 03:59:22 pb_: thanks!! Jun 13 04:01:07 * CoreDump|home boots a loop-image off of NFS and grins like a child Jun 13 04:01:21 pb_: Someone (you I gather) having taken on glibc for arm :) Jun 13 04:03:37 argh.... SMB is shit Jun 13 04:03:51 the shitsnizzle! Jun 13 04:05:40 RP: oh, heh. that isn't exactly news: I've been the arm glibc maintainer for the past eight years. Jun 13 04:06:29 pb_: ah :). Has the decision to axe glibc's arm support been reversed or was it never as rmk made out? Jun 13 04:07:43 RP: There never was any decision to "axe" glibc's arm support. Jun 13 04:08:02 As far as I know, all these conspiracy theories are just a figment of Russell's imagination. Jun 13 04:08:38 pb_: Right. He did seem to be adding 2 and 2 and coming up with 37... Jun 13 04:08:46 03pb 07 * r1.3566.1.1 10openembedded/packages/tslib/ (tslib_cvs.bb tslib/h1940/tslib.sh): add h1940 support to tslib, courtesy Arjan Schrijver Jun 13 04:09:10 pb_: Its nice to know its not the case though as I was wondering :) Jun 13 04:13:25 ah, were they brought on by the move to ports ? Jun 13 04:13:43 I found that a bit disturbing myself Jun 13 04:15:12 I think that was the trigger for this latest bout of hysteria, yeah. Jun 13 04:23:26 Hi all Jun 13 04:24:21 03CoreDump 07 * r1.3569 10openembedded/packages/altboot/altboot_20050614.bb: altboot: Update to latest version (support booting off of NFS) Jun 13 04:25:33 hi Dirk Jun 13 04:26:32 hi hrw|work Jun 13 04:27:02 how to say for smbfs that share mounted week ago does not exist now? Jun 13 04:27:08 in other way then reboot Jun 13 04:29:18 no idea Jun 13 04:29:37 maybe kill all processes accessing files on the share then umount it? Jun 13 04:30:11 lsof dies when trying to chec it Jun 13 04:30:35 heh Jun 13 04:30:44 fuser? Jun 13 04:31:05 nearly same Jun 13 04:31:34 dang Jun 13 04:31:57 uf.. finally umounted Jun 13 04:41:10 hey, I'm struggling with version control in OE here... Jun 13 04:41:35 why does OE/bitbake think '2.6.11-bk1' is newer than '2.6.11.11' ? Jun 13 04:41:39 later Jun 13 04:42:09 because '-' sorts higher than '.', at a guess. Jun 13 04:43:49 would that qualify as a bug? Jun 13 04:46:01 dunno. If you can produce a convincing explanation of why the current behaviour is wrong, it might. Jun 13 04:52:15 ~hail br1 Jun 13 04:52:19 * ibot bows down to br1 and chants, "I'M NOT WORTHY!!" Jun 13 04:59:37 wow! with glibc-2.3.5 I don't have to manually fix /usr/lib/libc.so and /usr/lib/libpthread.so anymore! Jun 13 05:00:31 (on armeb) Jun 13 05:01:06 jacques: rock! Jun 13 05:01:30 actually, I think libpthread.so should be OK even in 2.2.5 these days. Jun 13 05:01:40 iirc, I checked in a fix for that problem last week or so. Jun 13 05:01:57 are we using 2.3.5 on armel yet? Jun 13 05:02:00 no Jun 13 05:02:39 or, at least, I'm not. I guess some people might be. Jun 13 05:03:27 ok, that explains why it didn't get built Jun 13 05:04:56 14:01 hrw@home:wrt54$ bitbake nylon-image-mini Jun 13 05:05:35 lets test Jun 13 05:05:55 hrw|work: you have a serial cable ready? Jun 13 05:06:28 koen: not yet - first will try with openwrt images anyway Jun 13 05:07:03 koen: and bootloader on wrt54 accept tftp during first seconds after boot Jun 13 05:07:16 yeah Jun 13 05:07:25 that's how I tried to save my wap54g Jun 13 05:08:24 if that will fail then I will build serial Jun 13 05:09:09 anyway I'm tired with original firmware where I cannot check does router see my wifi card in zaurus or not Jun 13 05:10:34 its blueblackbox Jun 13 05:11:10 ok Jun 13 05:11:27 now every machine uses _jffs to specify jffs2 paramaters Jun 13 05:11:38 meters even Jun 13 05:12:06 wrt54oe needs work Jun 13 05:12:08 03koen 07 * r1.3566.2.1 10openembedded/conf/machine/jornada56x.conf: jornada56x.conf: guard jffs2 args with _jffs2 Jun 13 05:12:16 ERROR: The following variable(s) were not set: TARGET_OS Jun 13 05:12:17 Please set them directly, or choose a MACHINE or DISTRO that sets them. Jun 13 05:13:06 ;) Jun 13 05:15:12 pb_: can I talk you into adding a gtk+-native for those funky pixbuf binaries? Jun 13 05:56:27 igrth: sure Jun 13 05:57:29 (: Jun 13 05:59:32 igrth: http://article.gmane.org/gmane.comp.handhelds.openembedded/4416/match=guidelines+src+uri Jun 13 06:00:17 with a big thanks to hrw|work for summing it up Jun 13 06:00:26 thnx :) Jun 13 06:00:41 the latest incarnation of our .bb 'docs' are at http://ewi546.ewi.utwente.nl/tmp/d/ Jun 13 06:00:42 koen: no, that doesn't bear thinking about Jun 13 06:01:03 to build gtk+-native, you'd also need native versions of the whole X stack, freetype, fontconfig, and who knows what else. Jun 13 06:01:45 drat Jun 13 06:02:52 so parts of gpe will be FTBFS.. Jun 13 06:03:27 hrw|work: eh? Jun 13 06:04:22 pb_: some time ago I had failed building of some stuff.. it was not from gpe but it was one of gtk based apps Jun 13 06:05:11 ah. so, how does that lead you to suppose that "parts of gpe will be FTBFS"? Jun 13 06:06:00 if any of the gpe packages require gdk-pixbuf-csource at build time, they are just broken. Jun 13 06:06:00 I don't think that's the case, though. Jun 13 06:06:43 * koen gives the gimp the evil eye Jun 13 06:07:24 ops. it was gimp. Jun 13 06:07:25 sorry Jun 13 06:07:50 03 11:20< hrw|work> gimp fails on do_configure: | configure: error: Could not find gdk-pixbuf-csource in your PATH Jun 13 06:17:01 well, that's nice. xserver fails to build, and bitbake fails to paste the log to pastebin :/ Jun 13 06:19:42 bother, what is the command to unlock bk again? Jun 13 06:19:54 bk resolve? Jun 13 06:20:45 heh thanks Jun 13 06:21:04 I some wacking last week I had renamed a dir and this mucked things up .. Jun 13 06:27:53 03koen 07 * r1.3572 10openembedded/conf/machine/h1940.conf: h1940.conf: adjust xserver dp, couresy Arjan Schrijver Jun 13 06:27:54 gtk2 filereq is slow ;( Jun 13 06:31:16 koen: thanks! that was fast :) Jun 13 06:32:40 if you ever wanna really confiuse yourself, accidentally replace busybox with something else Jun 13 06:33:08 heh Jun 13 06:43:27 hmm gnu0config-native failing to patch Jun 13 06:43:38 SHould I change my CVSDATE? Jun 13 06:44:43 yep Jun 13 06:45:19 any dea what oe will use for source control after July 1? Jun 13 06:45:25 not yet Jun 13 06:45:46 deadline scheduling Jun 13 06:45:54 someone in a rush on july 1st will probably decide it Jun 13 06:48:13 xserver-kdrive won't compile... is that a known bug? Jun 13 06:48:35 it compiled 5 minutes ago Jun 13 06:48:52 well, 150 minutes to be honest Jun 13 06:49:21 Crofton|laptop: gnu-config 20050516 or newer Jun 13 06:49:23 if it is the shmfun thing, you'll need to build the new xextensions snapshot Jun 13 06:49:55 hrw|work, I just bumped my cvsdate to yesterday :) Jun 13 06:49:57 thanks Jun 13 06:50:48 koen: it's indeed a ShmFuncsPtr problem. how do I get the new xextensions? I started an hour ago with a completely empty tmp directory Jun 13 06:52:05 bitbake -b xextensions/xextensions_20050610.bb Jun 13 06:52:37 ok, thanks. how long before it will be default (i.e. preferred)? Jun 13 06:52:45 it should be default Jun 13 06:53:07 I'll have a look at this when this build run is finished Jun 13 06:53:23 well, it apparently isn't since it just build 'xextensions-1.0.1+cvs20050418-r0' by default Jun 13 06:58:25 yeah, the version string is wrong in the 20050610 one. Jun 13 06:58:36 PV = "0.0cvs${FIXEDCVSDATE}" Jun 13 06:58:39 that's no good Jun 13 06:59:12 that gets defined in the .bb Jun 13 06:59:25 i.e. it works for kdrive Jun 13 06:59:38 eh? Jun 13 06:59:56 check the xserver-kdrive_20050610.bb Jun 13 07:00:19 I don't understand what you're saying. Jun 13 07:00:22 'bitbake xextensions' -> 'NOTE: package xextensions-0.0cvs20050610: completed' Jun 13 07:01:14 pb_: FIXEDCVSDATE = "${@bb.data.getVar('FILE', d, 1).split('_')[-1].split('.')[0]}" Jun 13 07:02:03 yes, I know what FIXEDCVSDATE is. my point is that "0.0cvs${FIXEDCVSDATE}" is not a correct version string for xextension. Jun 13 07:02:31 doesn't that get expanded to 0.0cvs20050610 ? Jun 13 07:02:34 yes Jun 13 07:03:01 but 0.0cvs will sort lower than 1.0.1, let alone the older 1.0.1+cvs20050418 Jun 13 07:03:14 ah Jun 13 07:03:27 I thought 0.0 was the default versions for _cvs .bbs Jun 13 07:03:57 So did I but then I copied it out of one of the x .bb files :-/ Jun 13 07:03:59 and there's no DEFAULT_PREFERENCE in any of those .bb files, so the behaviour that Arjan is seeing is to be expected. Jun 13 07:05:03 I think the right thing to do is just take a copy of xextensions_1.0.1+cvs20050418.bb and change the date. Jun 13 07:05:10 PREFERRED_VERSION_xextensions ?= "0.0cvs20050610" Jun 13 07:05:26 that's in preferred-gpe-versions.inc Jun 13 07:05:52 right, but not everyone will be including that file. Jun 13 07:06:20 in any case, having versions go backwards is always bad news, since it will break package-based upgrades. Jun 13 07:06:30 that's what I thought too Jun 13 07:06:45 but I was overwhelmed with the 0.0cvs packages in OE Jun 13 07:07:33 if there's some overwhelmingly compelling reason why the version number needs to change from 1.0.1 to 0.0, you'd need to introduce an epoch. But in this case, I don't think there is any such reason. Jun 13 07:07:44 ok Jun 13 07:07:44 yeah, the "0.0cvs" packages are probably setting a bad example. Jun 13 07:08:48 it actually is 1.0.3 atm Jun 13 07:09:42 * koen watches a train scroll by Jun 13 07:09:57 I should learn to type 'ls' Jun 13 07:09:58 03RP 07 * r1.3559.1.1 10openembedded/packages/oprofile/ (4 files in 3 dirs): oprofile: Add a fix to correct the broken offset handling of the kernel image on arm Jun 13 07:11:04 mickey|lunch: when you go back from lunch can you build qte with w100 and then rebuild zroadmap? Jun 13 07:11:14 03koen 07 * r1.3572.1.1 10openembedded/packages/xextensions/xextensions_20050610.bb: xextensions_20050610.bb: change version to 1.0.3 Jun 13 07:11:23 | cp: cannot stat `/home/hrw/zaurus/oe/openembedded/packages/gcc/gcc-4.0.0/ldflags.patch': No such file or directory Jun 13 07:11:26 | Failed to import patch ldflags.patch Jun 13 07:11:29 NOTE: [16:10] Task failed: /home/hrw/zaurus/oe/build/wrt54/tmp/work/i686-linux/gcc-cross-4.0.0-r1/temp/log.do_patchcmd.23895 Jun 13 07:11:33 ~spell ambiguity Jun 13 07:11:35 'ambiguity' may be spelled correctly Jun 13 07:12:45 mickey|lunch: | /home/hrw/zaurus/oe/build/0602/tmp/cross/lib/gcc/arm-linux/3.4.3/../../../../arm-linux/bin/ld: warning: libaticore.so.1, needed by /home/hrw/zaurus/oe/build/0602/tmp/staging/arm-linux/qt2/lib/libqte.so, not found (try using -rpath or -rpath-link) Jun 13 07:13:23 strange... Jun 13 07:13:36 hrw|work: The atilib problem will be the w100 accel code :-/ Jun 13 07:13:42 tmp/staging/arm-linux/lib/libaticore.so.1.0.1: ELF 32-bit LSB shared object, ARM, version 1 (ARM), stripped Jun 13 07:14:01 it should not be stripped rather? Jun 13 07:14:24 hrw|work: Is there a link to that from libaticore.so.1 ? Jun 13 07:14:30 yes Jun 13 07:14:57 hmmm :-/ Jun 13 07:15:32 install -m 0755 libaticore.so.1.0.1 ${D}${libdir}/libaticore.so.1.0.1 Jun 13 07:15:45 and then it will be stripped? Jun 13 07:16:10 ignore me. its do_install not do_stage.. Jun 13 07:29:28 RP: what's the difference between aticore and aticore-oss and can the former be deleted? Jun 13 07:31:00 hrw|work: it works here. qte now depends on aticore-oss for c7x0, so that should've been built before Jun 13 07:31:21 mickeyl: and qte builds fine - zroadmap does not Jun 13 07:32:05 hmm Jun 13 07:32:26 why should we link to libraries which are sucked in by other libraries ? Jun 13 07:32:41 mickeyl: Yes, the former is broken in lots of ways and aticore-oss is superior in every way Jun 13 07:32:52 RP: very good. so no objection to remove it? Jun 13 07:33:00 mickeyl: none from me Jun 13 07:33:21 excellent. it'll be in the 'attic' if someone needs it Jun 13 07:33:34 hi mickey Jun 13 07:33:37 mickeyl: It won't actually work under 2.6 for a start ;-) Jun 13 07:34:26 RP: how's the w100 stuff going? Jun 13 07:34:50 hi pb_ Jun 13 07:34:51 koen: Haven't had a chance to work on it for the past month :-( Jun 13 07:35:11 i have some issues with gnome libs in oe Jun 13 07:35:21 RP: if I convince RMK not to break stuff anymore, would that help? Jun 13 07:35:23 i'll add 'em to the BT Jun 13 07:35:24 issues? oh dear. Jun 13 07:35:31 yeah, good plan Jun 13 07:35:35 koen: and akpm ;-) Jun 13 07:35:54 03mickeyl 07 * r1.3571.1.1 10openembedded/BitKeeper/deleted/.del-sharp-aticore.bb~19c6928d58961299: remove sharp-aticore. it has been superseded by sharp-aticore-oss Jun 13 07:36:35 someone with lots of coffee and time should upgrade gnome to 2.10 (or higher) in OE Jun 13 07:36:46 and add gnome-python Jun 13 07:36:59 One thing I'm impressed with though - after the removal of the preempt locks on the copypage functions, preempt *really* flies on the c7x0 Jun 13 07:37:09 RP: really? good to hear. Jun 13 07:37:23 RP: is that already in our rc5 build? Jun 13 07:37:31 RP: even on pxa250? Jun 13 07:37:32 mickeyl: Try a 2.6.12-rc5 kernel. Its really noticable Jun 13 07:37:36 RP: excellent. i'll have to give that a go. Jun 13 07:37:51 koen: even on a pxa250 Jun 13 07:38:05 rc5 has working pcmcia? Jun 13 07:38:35 koen as long as you umount /dev/hdax and cardctl eject before you pull the card out, its perfect ;-) Jun 13 07:38:51 as long as wifi works Jun 13 07:38:57 good morning CosmicPenguin Jun 13 07:39:02 morning Jun 13 07:39:21 I think it does. I loose track of when I last tested what which which versions Jun 13 07:40:12 * koen still wonders who's idea it was to put 32MB inthe corgi Jun 13 07:41:21 One thing that does worry me is why we now have to umount /dev/hdax Jun 13 07:41:51 I was blaming udev but I now have an image wityh udev 0.58 so its not that Jun 13 07:42:06 koen: they just upgraded poodle Jun 13 07:43:41 doesn't poodle have 64mb ram too? Jun 13 07:44:19 64flash, 32ram Jun 13 07:45:23 that's why it's the unloved step child Jun 13 07:46:02 idiots Jun 13 07:46:30 *nod* Jun 13 07:47:22 sharp engineer is often synonime to idiot Jun 13 07:49:52 get a microdrive and add swap :) Jun 13 07:50:02 I have swap on sd :) Jun 13 07:50:08 since i need wifi Jun 13 07:51:22 sd is flash? Jun 13 07:51:55 ya Jun 13 07:52:10 I would have thought swapping to flash would be "bad" Jun 13 07:53:13 not if ou think sd is the spawn of evil Jun 13 07:53:48 PROVIDES is space seperated, right? Jun 13 07:54:22 heh Jun 13 07:54:37 I have an old 8M CF card I will swap to one day to see how bad bad is Jun 13 07:56:24 koen: yep Jun 13 07:57:04 03koen 07 * r1.3574.1.1 10openembedded/packages/dropbear/dropbear_0.45.bb: dropbear_0.45.bb: PROVIDES='ssh sshd' Jun 13 08:01:12 * mickeyl wanders home Jun 13 08:02:29 03koen 07 * r1.3577 10openembedded/packages/hdparm/ (hdparm_5.8.bb hdparm-5.8/uclibc.patch): hdparm_5.8.bb: add uclibc patch , courtesy Bernhard Rosenkraenzer Jun 13 08:03:15 mickeyl: any updates on http://bugs.openembedded.org/show_bug.cgi?id=29 ? Jun 13 08:05:31 I built an opie-image yesterday without any failures... Jun 13 08:06:06 fails on, oops, I though it meant fails building (back to marking) Jun 13 08:06:08 03koen 07 * r1.3578 10openembedded/packages/gftp/gftp_2.0.18.bb: Rename: packages/gftp/gftp_2.0.18rc1.bb -> packages/gftp/gftp_2.0.18.bb Jun 13 08:08:21 koen: thanks Jun 13 08:08:41 koen: I was pulled off on other images, and I haven't built a X based image in eons, it seems Jun 13 08:09:25 CP|Laptop: wanna update gnome? Jun 13 08:09:30 heh Jun 13 08:09:38 * CP|Laptop gags Jun 13 08:09:47 thought so :) Jun 13 08:10:57 * koen taps foot Jun 13 08:11:01 come on sdl! Jun 13 08:11:53 I've been slowly updating packages as I find them Jun 13 08:12:19 but eventually, one of these days, one package is going to break, and all thousand packages will have to be updated accordingly Jun 13 08:12:22 I fear that day Jun 13 08:12:25 I stopped after the new gnome stuff crapped out on weird autofoo Jun 13 08:12:45 LIBFOO_AGE=1000 Jun 13 08:12:58 1000 > 8bits int Jun 13 08:14:03 03koen 07 * r1.3579 10openembedded/ (2 files in 2 dirs): libsdl-mixer_1.2.6.bb: Add it Jun 13 08:16:15 I have finally to re-arrange my development dirs... Jun 13 08:17:07 I have ~/src/ for desktop ~/src/php/ for websites ~/zaurus/oe/ for OE builds, ~/zaurus/oe/cvs/opie/ with checkout, now openwrt comes.. Jun 13 08:17:53 hrw|work: TMPDIR=/home/koen/OE/build/tmp/${DISTRO} Jun 13 08:18:11 works like a charm Jun 13 08:18:42 now, if we could just start sharing components... :) Jun 13 08:18:59 * CP|Laptop builds six i486 images a night Jun 13 08:19:54 * koen is busy with a 12 machines x 4 images build Jun 13 08:20:42 koen: drop the opie builds ;) Jun 13 08:21:04 I wish :) Jun 13 08:25:45 koen: ok. but I think about more general development dir Jun 13 08:32:25 hrw|work: If you solve the problem, let me know as I'm having similar problems :) Jun 13 08:33:35 RP: currently I will go with ~/devel/openwrt/ ~/devel/php/ ~/devel/OE ~/devel/opie probably Jun 13 08:36:51 mickeyl: could you or zecke have a look at the bitbake bugs (58-65 Jun 13 08:36:55 )? Jun 13 08:39:46 there, that's enought bugzilla spam for now Jun 13 08:39:54 cu Jun 13 08:40:01 cu Jun 13 08:41:03 heh Jun 13 08:41:14 neon should probably stop distributiong .la files Jun 13 08:41:46 Resolved FIXED (CLUELESS) Jun 13 08:45:14 CP|Laptop: heh Jun 13 08:47:07 * kergoth yawns Jun 13 08:48:08 hmm, I had a problem with libstdc++ doing that yesterday I think Jun 13 08:48:17 good morning kergoth Jun 13 08:48:43 yeah, /usr/lib/libstdc++.la has wrong paths in it Jun 13 08:48:51 hey Jun 13 08:49:00 I hate it when that happens Jun 13 08:49:13 as does /usr/lib/libsupc++.la Jun 13 08:49:31 mm, that's rather sad. Jun 13 08:49:44 I guess we need some kind of libtool-munging pass to straighten it out. Jun 13 08:50:13 pb_, 2.3.5 armeb is looking very good so far - I am building perl, but by the time it's done testing you'll probably be asleep :-) Jun 13 08:50:30 jacques: excellent Jun 13 08:50:39 heh, yeah. well, you can alert me to the results tomorrow :-) Jun 13 08:50:45 yep :-) Jun 13 08:51:18 I consider perl configure/build/test to be a pretty good indicator of libc functionality Jun 13 08:51:29 * kergoth hooks up his old spaceorb 360 controller and plays around Jun 13 08:52:29 hey, I had some of those :-) Jun 13 08:53:29 there's a 2k/xp spaceorb driver done by a user, including a new customizer app to configure the axis' and stuff. :) /me just got his ass kicked in half life trying to use it Jun 13 08:54:32 heh, I think the only thing they were really good for was descent Jun 13 08:54:58 they were perfect for descent. pretty good for first person shooters, but it always had a steep learning curve Jun 13 08:55:24 um descent Jun 13 08:55:25 for fps give me a keyboard and mouse Jun 13 08:55:36 i cant get the hang of console fps control Jun 13 08:55:41 descent... Jun 13 08:56:45 descent was a good game Jun 13 08:57:15 I know a guy still playing D3 Jun 13 08:57:17 mmm... descent Jun 13 08:57:22 descent was the main use of our school network Jun 13 08:57:36 486sx with ipx over bnc Jun 13 08:57:47 I never had DSL when I played descent Jun 13 08:57:56 that limited me badly Jun 13 08:58:01 Did anybody have DSL when they played descent? Jun 13 08:58:13 I tried playing d3 online a few times, but the other players were just amazing Jun 13 08:58:17 heh, my freind did Jun 13 08:58:20 yeah Jun 13 08:58:23 It was tough Jun 13 09:00:02 i remember loving descent Jun 13 09:00:39 i remember when i finally learned to turn off the snap-to-floor flight thing Jun 13 09:02:26 heh, we need to resurrect old games like that ... Jun 13 09:03:10 there was a linux port of descent Jun 13 09:03:19 i don't think it went anywhere though Jun 13 09:04:32 http://www.lokigames.com/products/descent3/ Jun 13 09:04:35 * Crofton|laptop weeps Jun 13 09:05:10 too bad parsec was basically stillborn too Jun 13 09:07:29 freespace also rocked Jun 13 09:26:08 :) Jun 13 09:26:17 anyone with hostusb working awake? Jun 13 09:26:23 i'm very afraid that i've fried my zaurus Jun 13 09:28:03 * koen feels a "topic, 3rd item" coming Jun 13 09:28:14 OE != OZ Jun 13 09:28:29 i know Jun 13 09:28:39 well, how would i test hostusb if it's fried or not Jun 13 09:28:50 that's more of a kernel / OE question non? Jun 13 09:28:56 no Jun 13 09:29:07 it has nothing to do with *building* stuff Jun 13 09:29:17 hm Jun 13 09:29:17 evening Jun 13 09:29:38 o-well Jun 13 09:52:47 koen|supper: okay... http://www.oesf.org/forums/index.php?showtopic=12943&pid=83398&mode=threaded&show=&st=&#entry83398 Jun 13 09:52:56 who is Lurker? Jun 13 10:02:15 jnc: I repeat OE != OZ Jun 13 10:02:36 plugging usb cables has NOTHING to do with building stuff Jun 13 10:03:13 03koen 07 * r1.3580 10openembedded/packages/maemo/osso-af-utils_0.4.bb: osso-af-utils_0.4.bb: remove staging Jun 13 10:04:27 koen|supper: o_O Jun 13 10:05:06 i'm not even running OZ Jun 13 10:05:09 what's your problem buddy Jun 13 10:12:17 jnc: do you ask the rpm people how to test a networkcable? Jun 13 10:25:03 koen: maybe the rpm people are friendly to share ideas about where to find that information Jun 13 10:25:27 * jnc whispers to santa "lump of coal for that one *points*" Jun 13 10:26:10 the manual, or the sharp support website would be my guesses Jun 13 10:26:31 Lurker seems to know, his post is on OE forums Jun 13 10:26:39 you know if Lurker frequents IRC? Jun 13 10:26:42 its trolltime? Jun 13 10:26:50 those forums have nothing todo with this Jun 13 10:27:04 maybe a small bit Jun 13 10:27:23 ~seen lurker Jun 13 10:27:24 lurker <~anoland@216.17.243.15> was last seen on IRC in channel #familiar, 513d 23h 41m 32s ago, saying: 'Everybody sleeping?'. Jun 13 10:28:18 we avoid shuffling arbitrary binaries in images like the plague here Jun 13 10:28:30 the oesf people seem to have a fetish on that Jun 13 10:28:46 "look at me, I made my own ROM" Jun 13 10:28:51 :) Jun 13 10:28:59 koen *g* Jun 13 10:29:02 who is running oesf btw? Jun 13 10:29:07 * koen smacks those people with a dictonary Jun 13 10:29:12 offroadgeek Jun 13 10:29:21 a *read only* dictionary Jun 13 10:29:34 hmm Jun 13 10:30:07 it's kinda sad the forum is a blackhole for fixes and documentation Jun 13 10:30:28 and people call you an "idiot" if you suggest using the bugtrackers Jun 13 10:31:25 that's not quite true Jun 13 10:31:33 yah.. there's a lot of valuable knowledge in there and no sane way to get to that Jun 13 10:32:46 Get people to add to a wiki, that was the plan Jun 13 10:33:39 well.. I'm mostly referring to bug reports and proposed fixes or even patches Jun 13 10:35:05 pb_, All tests successful. Jun 13 10:35:11 perl tests out 100% Jun 13 10:35:13 jacques: awesome Jun 13 10:35:25 I'm happy :-) Jun 13 10:35:48 one (1!) person posted a bug to the gpe bugzilla Jun 13 10:36:27 the other gazillion just go on complaining how stuff never gets fixed Jun 13 10:37:15 03koen 07 * r1.3581 10openembedded/packages/flite/flite_1.2.bb: flite_1.2.bb: add staging, courtesy Patrick Ohly Jun 13 10:37:22 bye everyone Jun 13 10:37:26 cu Jun 13 10:38:26 pb the simpad is booting with glibc 2.3.5 Jun 13 10:38:40 koen: since when is "make install" accepted for do_stage() ? Jun 13 10:38:50 pb didt test somemore becaus zecke forgot some function in the mq200 driver Jun 13 10:39:17 no one have this error when building something: whereas xxxxx uses software FP Jun 13 10:39:38 reenoo_: autotools_stage_includes does the same Jun 13 10:39:59 woglinde: great Jun 13 10:40:10 reenoo_: that bbclass does the same but uses DESTDIR instead of INSTALLDIR Jun 13 10:41:14 will armv5 apps run with an armv4 glibc? Jun 13 10:43:47 koen: I'm still not convinced that's a good idea. the idea behind do_stage() was to keep access to staging as restricted as possible Jun 13 10:44:51 koen: i.e. oe_libinstall instead of some random make install which might trash half of staging Jun 13 10:46:02 reenoo_: it probably has more to do with rapid prototyping then anything Jun 13 10:46:21 If you want to get something working now, make install is a hell of a lot better then copying over a hundred header files by hand.. :) Jun 13 10:46:38 autotools_stage_includes() doesn't "make install" directly into staging. It installs into ${STAGING_TEMP}, then pulls out the files it wants from there and copies them to the real staging area. Jun 13 10:47:55 pb_: that'd be fine. (I didn't actually look it up in autotools.bbclass right now) Jun 13 10:49:47 CosmicPenguin: there's nothing wrong with rapid prototyping in your local tree. I just don't see a reason why it should be pushed into the OE repo. Jun 13 10:50:08 i agree with reenoo_, better not pollute staging area Jun 13 10:50:20 let me check what files need to be installed Jun 13 10:54:58 mickeyl: thanks Jun 13 10:56:15 anyone on the armv4 vs armv5 question? Jun 13 10:58:25 yes, you can mix armv4 and armv5. Jun 13 10:59:50 eeek Jun 13 10:59:52 an ICE Jun 13 11:00:03 | /local/pkg/oe/c7x0/tmp/cross/lib/gcc/arm-linux/3.4.4/../../../../arm-linux/bin/ld: BFD 2.15.94.0.1 20041121 internal error, aborting at /local/pkg/oe/c7x0/tmp/work/binutils-cross-2.15.94.0.1-r0/binutils-2.15.94.0.1/bfd/cache.c line 495 in bfd_cache_lookup_worker Jun 13 11:00:04 * koen blames softfloat Jun 13 11:00:05 * chouimat is away: FOOD Jun 13 11:00:26 ah more an IBE Jun 13 11:00:30 heh Jun 13 11:00:38 koen whats wrong with softfloat? Jun 13 11:00:51 it used to ICE on gnuplot Jun 13 11:01:17 hm whats an ICE? Jun 13 11:01:27 internal compiler error Jun 13 11:01:44 oh silly me Jun 13 11:01:46 or in car entertainment Jun 13 11:01:53 intrusion countermeasures electronics Jun 13 11:02:01 if you;re into cyberpunk Jun 13 11:02:18 or the frozen water stuff written in capitals Jun 13 11:02:51 koen *g* Jun 13 11:03:05 koen some train in germany haha Jun 13 11:05:31 that too Jun 13 11:08:28 <[cc]smart> seems my suggestions for changes usually reach nobody with the time to weve them in. do you guys think it's worthwhile for those rare cases where i see myself how to do it, to be able to apply it myself and upload that ? Jun 13 11:09:16 what changes are we talking about ? Jun 13 11:09:36 <[cc]smart> e.g. in ntp i'd add local clock to ntp.conf Jun 13 11:10:23 <[cc]smart> whatever. key is i see no possibility to reliably feed in sth. Jun 13 11:10:59 <[cc]smart> if the bugreporting tool would actually be used... i'd see more sense in that way of doing it. Jun 13 11:11:03 i think that's a problem of the mailing list. would it be ok for you to submit some patches via bugzilla ? Jun 13 11:11:06 <[cc]smart> but it isn't Jun 13 11:11:11 bugzilla is heavily used Jun 13 11:11:16 the oe one, at least Jun 13 11:11:21 (we are talking about oe, aren't we?) Jun 13 11:11:45 <[cc]smart> i'm primarily talking about openslug, but that boils down to the same thing. no ? Jun 13 11:11:51 more or less, yeah Jun 13 11:11:57 we share the metadata Jun 13 11:12:17 <[cc]smart> so do oe devs use openslug bugzilla ? Jun 13 11:12:23 hmm Jun 13 11:12:28 i was talking about bugs.openembedded.org Jun 13 11:12:41 that's the preferred way to submit patches Jun 13 11:12:48 they don't get lost that way Jun 13 11:13:14 <[cc]smart> i'll check what it actually was i checked into andd compare to the one you just mentionned. Jun 13 11:13:15 and stuff like ntp really is not openslug specific so the oe bugtracker seems like the natural place Jun 13 11:13:20 k Jun 13 11:14:04 <[cc]smart> in any case, whatever the others are, they should get removed. they mean nothing than data loss if not used anyways... Jun 13 11:19:44 right Jun 13 11:20:44 <[cc]smart> hmm... at least historically i have to disagree with the usage of bugs.openembedded.org Jun 13 11:20:54 <[cc]smart> it's exactly that place i posted the bugs Jun 13 11:21:04 <[cc]smart> and at the time the story roughly went like this: Jun 13 11:21:13 <[cc]smart> < reported the bug Jun 13 11:21:31 <[cc]smart> --- time --- ( no problem with that, free things need time) Jun 13 11:21:42 the bugzilla or the RT? Jun 13 11:21:47 <[cc]smart> ... some more time, i notied somebody fixed the bug, without knowing i reported that Jun 13 11:22:08 we dumped RT and lost all bugs Jun 13 11:22:12 <[cc]smart> i told OE irc that the specific bug got fixed. the actual person who fixed it was online Jun 13 11:22:29 <[cc]smart> he tried to cose it but wasn't able to due to access rights Jun 13 11:22:35 <[cc]smart> nobody knew how to close a bug Jun 13 11:22:37 <[cc]smart> story end Jun 13 11:22:40 sounds like RT Jun 13 11:22:46 we have bugzilla now :) Jun 13 11:22:55 <[cc]smart> could be the reason Jun 13 11:22:57 <[cc]smart> :) Jun 13 11:23:15 the "quick report" featue of RT was wonderfull Jun 13 11:23:22 <[cc]smart> that's what i wanted to mention now. the two reports are gone, ids dont exist anymore .) Jun 13 11:23:26 I've never sen soo many empty bugreports Jun 13 11:23:49 seen even Jun 13 11:24:12 [cc]smart: I went through all open bugs in bugzilla today Jun 13 11:27:20 <[cc]smart> nice... what is 'low' priority. P1 or P5 ? Jun 13 11:31:27 no idea on that Jun 13 11:32:04 [cc]smart: p5 Jun 13 11:32:12 later all Jun 13 12:08:59 mickeyl: I dont see the tie-in for QWS_DISPLAY=W100 in that aticore patch. did you forget to diff qwsmouse_qws? Jun 13 12:09:59 * chouimat is back. Jun 13 12:10:38 ... Jun 13 12:10:43 there is an aticore driver now? Jun 13 12:10:54 who do I have to kiss? Jun 13 12:11:00 ggilbert: sirfred Jun 13 12:11:12 pucker up. Jun 13 12:11:25 and RP Jun 13 12:11:31 RP: You too Jun 13 12:14:18 aticore driver for what? X ? Jun 13 12:14:25 qte Jun 13 12:14:28 oh Jun 13 12:14:35 for the fb fetishists Jun 13 12:14:41 w100 chip Jun 13 12:14:51 Spyro: mallum promissed to write a kdrive driver for it Jun 13 12:16:23 koen: bugger. I was going to do that Jun 13 12:16:31 koen: has he done anything yet ? Jun 13 12:17:38 Spyro: not yet Jun 13 12:17:50 and I think he'll be happy with all the help he can get Jun 13 12:18:08 koen: unless hes desperate for that project, its on my todo anyway for once I get the eseries actually running properly Jun 13 12:18:11 Spyro: feel free ;-) Jun 13 12:18:14 How is performance on it? Jun 13 12:18:21 any improvement over straight w100fb? Jun 13 12:18:55 ggilbert: It uses accelerated blits, lines and rects Jun 13 12:18:57 so yes Jun 13 12:19:06 cool Jun 13 12:19:16 I'll have to try building it this week Jun 13 12:19:38 ggilbert: I think Sirfred said it was slower at some things and needed some tweaking Jun 13 12:20:20 ggilbert: but its a definite move the rhe right direction :) Jun 13 12:20:27 yeah Jun 13 12:20:33 details can always be ironed out Jun 13 12:21:30 hi Jun 13 12:21:43 Spyro: you fix for devmem2 pushed Jun 13 12:21:53 hrw|tv: cheers Jun 13 12:22:01 hrw|tv: which fix was that? Jun 13 12:22:26 S="${WORKDIR}" Jun 13 12:22:30 ohhh Jun 13 12:22:34 for the .bb Jun 13 12:22:50 was gonna say, I didnt remember fixing devmem2 itself :) Jun 13 12:23:06 I only submitted it, I think it was koen actually provided the fix Jun 13 12:23:28 RP: no excuses now, include devmem2 in your bb builds today :) Jun 13 12:23:42 Spyro: perhaps - but mail was from you Jun 13 12:23:52 heh :) Jun 13 12:24:21 Spyro: I've had that build for ages ;-) Jun 13 12:24:27 :) Jun 13 12:27:09 2.6.12-rc5-mm2-r1 building... Jun 13 12:28:03 RP: you know that patch you use for X ? Jun 13 12:28:12 RP: is all it does to disable apm ? Jun 13 12:28:26 Spyro: yes Jun 13 12:28:29 RP: I thought you said it did some vt switching or something as well ? Jun 13 12:28:50 Spyro: No, by disabling apm you stop that Jun 13 12:29:25 RP: do you do anything other than your patch to make it work ? Jun 13 12:29:57 Spyro: rm /dev/apm_bios? :) Jun 13 12:30:18 tried that Jun 13 12:30:35 no idea then Jun 13 12:31:24 hm. do you know if its possible to suspend any one given task using some command? Jun 13 12:31:36 apm -s ? Jun 13 12:31:51 koen: one task, not all of them! Jun 13 12:32:51 ah Jun 13 12:33:53 I'm not sure its possible Jun 13 12:34:03 RP: thats a glaring ommission... Jun 13 12:34:05 :) Jun 13 12:34:20 I've been through the suspend/resume code before and there is no such per process hook Jun 13 12:34:30 I'm also not sure it would be very useful Jun 13 12:34:30 I know its *possible* if you are the parent of a task (eg. you can halt a process int he shell with ctrl-z Jun 13 12:34:51 That's different to freezing it though, isn't it? Jun 13 12:35:59 only insomuch as its not via the apm mechanism. the task is still actually frozen... Jun 13 12:36:17 doesnt even receive sigkills until its woken Jun 13 12:36:28 I noticed killefiz came in here 42 days ago. anyone keep in touch with him? Jun 13 12:38:07 X is WAY to easy to send D-state Jun 13 12:39:27 subdue: iirc killefiz is/was replaced by elsix Jun 13 12:39:31 how does one cause the display to switch from one vt to another ? Jun 13 12:39:37 chvt Jun 13 12:39:40 cheers Jun 13 12:40:00 yes, I know. I've been trying work with elsix Jun 13 12:40:31 bah. X sucks :) Jun 13 12:40:51 yeah, let's all dump X and go play outside Jun 13 12:40:54 it's been painful Jun 13 12:41:06 koen: dont make fun ;-) Jun 13 12:41:40 * Spyro replaced the fanbelt today on his car Jun 13 12:41:51 it must have had what, 40% of its teeth left :) Jun 13 12:42:04 brought mine to the body shop to get the passenger side repainted Jun 13 12:42:29 ggilbert: didn't you have a cool mustang? Jun 13 12:42:33 yeah Jun 13 12:42:38 bon giorno Pigi Jun 13 12:42:42 neighbor ran into it while he was parking Jun 13 12:42:47 * koen envious of ggilbert Jun 13 12:42:49 so i need to get a dent removed and fix the paint Jun 13 12:42:51 also repaird the throttle cable mount, tighteded throttle cable, tightened handbrake from near useless to working... Jun 13 12:43:02 Hi koen Jun 13 12:43:02 (and people owning cars in general) Jun 13 12:43:06 Ciao all Jun 13 12:43:08 hehe Jun 13 12:43:15 if I could get away with not having one, I'd love it Jun 13 12:43:15 * Spyro feels all included now :) Jun 13 12:43:28 they are generally more trouble than they are worth :) Jun 13 12:43:44 * Spyro pretends to own the merc outside the house instead of the daihatsu 5 meters down the road :) Jun 13 12:43:52 hehe Jun 13 12:44:11 actually I love my little Charade. its an '88 and still runs fine Jun 13 12:45:15 allegedly capable of 70mpg and up to 100mpg in town driving... Jun 13 12:45:27 driving a 2002 or so jeep grand cherokee. Man I miss mine already Jun 13 12:45:38 Spyro: yum Jun 13 12:45:55 Spyro: afaik the o/s ati lib needs some aditions before an X driver can be written Jun 13 12:45:57 ggilbert: she goes too - not bad at all in low gear Jun 13 12:46:14 mallum: I was going to start from scratch and not use the atilib stuff Jun 13 12:46:16 luckilly we have "relatively" cheap gass Jun 13 12:46:24 and ours is the most expensive in the country Jun 13 12:46:36 ggilbert: heh, luckily for me my cars a turbo diesel :) Jun 13 12:46:51 diesel cheaper over there? Jun 13 12:46:56 I know gas is fucking expensive Jun 13 12:47:01 koen, your italian is improving btw Jun 13 12:47:03 no, but turbo diesels do better MPG :) Jun 13 12:47:08 :) Jun 13 12:47:10 diesel is less fuckiing expensive Jun 13 12:47:11 Spyro: um... Jun 13 12:47:18 Pigi: :) Jun 13 12:47:18 it kind of varies here Jun 13 12:47:47 just curious, do gas prices in europe fluctuate daily like they do here? Jun 13 12:47:58 ggilbert: yeah Jun 13 12:48:05 fun Jun 13 12:48:28 ggilbert: mind you I've only had to fill her once... ~250 miles ago :) Jun 13 12:48:34 ggilbert, not toomuch here in Italy, but they are high, very high Jun 13 12:48:49 I can't imagine going to the supermarket every week and not knowing how much a bottle of water would cost Jun 13 12:48:52 or a loaf of bread Jun 13 12:48:59 from one day to the next Jun 13 12:49:11 ggilbert: welcome to my world :) Jun 13 12:49:17 ggilbert: I cant remember where I live :) Jun 13 12:49:22 mallum: ? Jun 13 12:49:41 Spyro: you need the atilib to know how to access the accell funcs of the chip Jun 13 12:50:09 mallum: I was going to start with basic fb and add functions one by one as I go through atilib Jun 13 12:50:22 I already know how to do one or two of the accel functions Jun 13 12:50:58 Spyro: but it makes more sense to have it use atilib and share the code Jun 13 12:51:23 mallum: I'll have to investigate. ATIlib has changed somewhat since I last looked at X drivers Jun 13 12:51:54 mallum: does atilib actually use the chips ringbuffer system properly ? Jun 13 12:52:12 Spyro: yeah I know RP was looking at it Jun 13 12:52:34 anyhow, its on my todo :) Jun 13 12:52:52 just as soon as I get X to sus/res :) Jun 13 12:53:14 Spyro: just kill the apm code Jun 13 12:53:37 mallum: easier said than done. I've tried that and still have problems Jun 13 12:54:02 Spyro: hmm, RP is the guy to speak to about this Jun 13 12:54:29 I went for a drive earlier and the MG scared me to death. I think new tyres are needed :-/ Jun 13 12:54:49 Spyro: we really should all try and co-ordinate X work for c7x0's Jun 13 12:55:27 RP: I just got my spare de-punctured :) Jun 13 12:55:57 Spyro: When your car's backend steps two foot sideways coming off a roundabout, you start to worry... Jun 13 12:56:05 RP: heh Jun 13 12:56:20 and I really do mean two foot - I'm not exagerating... Jun 13 12:56:22 That's rear wheel drive for you :-/ Jun 13 12:56:24 heh :) Jun 13 12:56:57 ok that was annoying. Ran X, changed VT, suspended, resumed, X dead Jun 13 12:57:06 mallum, Spyro: An X driver could be written against atilib as it stands. I'd much prefer to finish decoding the important bits of it and write something independent though Jun 13 12:57:29 RP: that was my plan anyway :-) Jun 13 12:59:20 Spyro: Ideally, I'd like to throw a lot of it into w100fb... Jun 13 12:59:47 RP: yeah but I dont really want w100fb loading microcode... Jun 13 13:00:39 Good night Jun 13 13:01:10 hello and 'night SirFred Jun 13 13:01:16 mickey|tv: ping Jun 13 13:01:28 Spyro: We request it using the existing firmware mechanism Jun 13 13:01:29 doesnt busybox provide nohup ? Jun 13 13:01:44 RP: not much use when the root FS isnt mounted... Jun 13 13:03:05 Spyro: So you fall back to software accel until you get that... Jun 13 13:03:40 does the fb layer allow you to hotswap the accel functions ? Jun 13 13:05:20 Spyro: Yes, you just do a fallthrough Jun 13 13:06:05 I'm getting this error trying to build gpe-image: "sh: ipkg-build: command not found". Any ideas on how to fix it? Jun 13 13:06:21 hey this is interesting Jun 13 13:06:31 X dies if I run chvt Jun 13 13:06:36 no suspend required Jun 13 13:06:45 mreimer: what distro? Jun 13 13:06:52 Spyro: It doesn't do that here... Jun 13 13:07:03 RP: what build ? Jun 13 13:07:05 hrw|tv: DISTRO="familiar-0.8.3" Jun 13 13:07:22 mreimer: hmm.. ipkg-utils-native was build? Jun 13 13:07:24 hi Zero_Chaos Jun 13 13:07:27 zecke: hi Jun 13 13:07:33 hey Jun 13 13:07:43 hrw|tv: supposedly: "ipkg-utils-native-1.6cvs20050613: started" and "completed" Jun 13 13:07:46 * zecke lies in bed with the laptop on the lap... Jun 13 13:07:50 Spyro: openzaurus-3.5.4 c7x0 Jun 13 13:07:57 RP: I meant X Jun 13 13:08:03 RP: mines CVS 0610 Jun 13 13:08:17 mreimer: hmm.. Jun 13 13:08:46 ARGH.... ipkg-collateral replace ipkg.conf ;(( Jun 13 13:08:50 Spyro: I have a CVS 0610 image ready to test but haven't got around to that yet Jun 13 13:09:04 RP: is our current one softfloat or hard ? Jun 13 13:09:28 Actually, I've been running softfloat binaries on a hardfloat image today :-/ Jun 13 13:09:38 How has that worked?! Jun 13 13:09:56 RP: can you send me your Xfbdev binary ? Jun 13 13:10:04 Spyro: I nearly always use softfloat Jun 13 13:10:29 Spyro: softfloat or hardfloat? Jun 13 13:10:40 hard preferred Jun 13 13:12:05 Spyro: http://www.rpsys.net/openzaurus/temp/xserver-kdrive-fbdev_0.0cvs20050612-r10_armv5te.ipk Jun 13 13:12:20 As you'll note, I used 0612 for some reason Jun 13 13:12:38 RP: 0612 has extra fixes? Jun 13 13:12:50 koen: No, I just messed the dates up :) Jun 13 13:12:52 feel free to adjust the cvsdate (bk mv it) Jun 13 13:13:05 RP: I built mine when 612 wasnt available... Jun 13 13:13:16 it's hardcoded to 0610 Jun 13 13:13:43 hrw|tv: what's that mean, about ipkg-collateral? Jun 13 13:14:00 koen: I'll try and talk mallum into adding the apm fix next ;-) Jun 13 13:25:01 RP: your binary does the same thing Jun 13 13:25:45 RP: can you try booting in to runlevel 2, running Xfbdev (nothing extra) and then chvt 2 and chvt 1 Jun 13 13:25:50 koen: where did you add md5sum to? Jun 13 13:25:53 see if X survives? Jun 13 13:25:55 QtE 2.3.10? Jun 13 13:26:19 Spyro: I need to install that image first and I have stuff on device that will cause issues with Jun 13 13:26:42 RP: can you try with your current image ? Jun 13 13:26:55 Spyro: Its an opie image Jun 13 13:26:59 oh Jun 13 13:27:02 ah well Jun 13 13:27:14 hm. perhaps I'll run gdb on it :) Jun 13 13:28:43 interesting Jun 13 13:28:49 segfault in ts_close Jun 13 13:29:26 Spyro, meet mallum :) Jun 13 13:29:36 RP: heh :) Jun 13 13:29:41 is this one known about? Jun 13 13:29:48 :) Jun 13 13:29:53 is this with latest CVS ? Jun 13 13:29:53 basically a chvt is causing a segfault in ts_close Jun 13 13:30:06 mallum: 0611 from RP Jun 13 13:30:10 mallum: yes - its upon chvt Jun 13 13:30:10 and 0610 Jun 13 13:30:21 0612 actually Jun 13 13:30:26 RP: :) Jun 13 13:30:41 RP: ok it works fine till you chvt ? Jun 13 13:31:01 mallum: yes, although it doesnt like me closing the last client much (sometimes dies) Jun 13 13:31:10 mreimer: ipkg-collateral is a package which contain ipkg.conf Jun 13 13:31:11 I bet its whats nuking it on suspend too Jun 13 13:31:29 can I suppress it using tslib with a command line switch ? Jun 13 13:31:29 mallum: I've not tested this yet. I'll do so as a priority though Jun 13 13:32:39 o.. nearly forgot... Jun 13 13:32:55 RP: I mapped Fn+Up/Down as PgUp/PgDown Jun 13 13:33:49 hrw|tv: Can we add a keymappings document to the documentation on openzaurus.org? Jun 13 13:33:58 Add info about this, plus sysrq etc Jun 13 13:34:08 RP: we can Jun 13 13:34:09 and ctrl+alt mappings Jun 13 13:34:19 wierd, kdrive tslib.c could really do with a rewrite. Jun 13 13:34:23 ofcourse - someone have to start it Jun 13 13:35:26 aha Jun 13 13:35:39 if I disable tslib (unset TSLIB_* in env) it will behave Jun 13 13:36:28 it even survives sus/res Jun 13 13:37:11 Spyro: the fix should be pretty trivial in X too I would have suspected Jun 13 13:37:28 hrw|tv: pong Jun 13 13:37:51 mickey|tv: run opie with W100:0 and then "gfx rendering" test in opie-sysinfo. rect blits are not synced correctly - some of them appear after 'opie logo' Jun 13 13:38:10 mallum: heres the offending code Jun 13 13:38:16 int ts_close(struct tsdev *ts) Jun 13 13:38:16 { Jun 13 13:38:16 int ret; Jun 13 13:38:16 ret = close(ts->fd); Jun 13 13:38:16 free(ts); Jun 13 13:38:17 return ret; Jun 13 13:38:19 } Jun 13 13:38:19 zap: ping Jun 13 13:38:25 hrw|tv: possible. I didn't test yet. sirFred says it's under construction and prone to emit some gfx errors yet Jun 13 13:38:32 ts = NULL Jun 13 13:38:41 hrw|tv: is it faster ? Jun 13 13:38:49 didn't Xora have a patch for that? Jun 13 13:39:03 mickeyl: 76567 gops/sec - didnt remember how much was before Jun 13 13:39:09 mallum: I would think it'd be good if tslib didnt use the pointer unconditionally :) Jun 13 13:39:17 ljp: i don't know what binding you are missing. it's listed in qgfx - no need to specify it in qmouse Jun 13 13:39:24 hrw|tv: uhm Jun 13 13:39:31 hrw|tv: i think it was ... like 10000 or so :) Jun 13 13:39:33 if not even less Jun 13 13:39:35 heh Jun 13 13:39:40 mickeyl: 932,33 now Jun 13 13:40:14 mickeyl: circles are drawed slower - some slower and I could even count them Jun 13 13:40:19 Spyro: This was what the apm patch was supposed to work around Jun 13 13:40:24 Spyro: and no doubt TslibDisable() above that ( via xserver/hw/kdrive/linux/tslib.c ) Jun 13 13:40:28 heh.. ya, I thought about what I said after i had a cup of coffe and it didnt make much sence Jun 13 13:40:30 Spyro: X was closing the ts device twice over Jun 13 13:40:47 RP: ah :) Jun 13 13:40:51 ljp: hey Jun 13 13:40:58 ljp: is Qt/E rc1 available yet? Jun 13 13:40:59 hi zecke Jun 13 13:41:05 RP: well, I applied the APM patch and it still fucked up :) Jun 13 13:41:07 probably check for NULL in TslibDisable() too :) Jun 13 13:41:11 snapshot Jun 13 13:41:17 I'll fix tslib Jun 13 13:41:32 mallum: you think tslib should print a warning to stderr? Jun 13 13:41:42 ljp: what about my Qt/E tslib patch? Should I send formal verification to show its correctnes? Jun 13 13:42:20 Spyro: I don't think code should be freeing a null device in the first place really Jun 13 13:42:38 didnt I apply that? Jun 13 13:42:50 hmm Jun 13 13:42:50 RP: yeah but tslib isnt exactly robust in that regard is it? Jun 13 13:42:56 | /bin/sh: /local/pkg/oe/c7x0/tmp/staging/i686-linux/bin/glib-mkenums: /usr/bin/perl5: bad interpreter: file not found Jun 13 13:42:58 wtf ? Jun 13 13:43:06 while building pango Jun 13 13:43:19 hrw|tv: I think sirFred|tv said somethings were faster with accel, some slower. We need to work out which cases we should use the hwaccel... Jun 13 13:43:22 ljp: you said it gave a wrong pointercal Jun 13 13:43:36 mickeyl: doesn't glib hve perl-native in depends? Jun 13 13:43:50 RP: ok Jun 13 13:43:56 i'll double check Iapplied it when I get into work Jun 13 13:44:00 koen: it has, however see here: Jun 13 13:44:08 more /local/pkg/oe/c7x0/tmp/staging/i686-linux/bin/glib-mkenums Jun 13 13:44:13 #!/usr/bin/perl5 -w Jun 13 13:44:16 eek :) Jun 13 13:44:18 heh Jun 13 13:44:20 ~lart absolute paths Jun 13 13:44:21 * ibot readies the nuke launcher and fires some rounds at absolute paths Jun 13 13:44:23 Does anyone know offhand whether sourceforge anoncvs is delayed compared to the actual edvel repositry? Jun 13 13:44:32 it is Jun 13 13:44:37 approx 5 hours Jun 13 13:44:45 argh Jun 13 13:44:51 5 to 24 hours actually Jun 13 13:44:53 bitbake just nuked my fix Jun 13 13:45:07 koen: should i add it into the BT ? Jun 13 13:45:07 is it possible to get bitbake to rebuild without re-fetching/unpacking ? Jun 13 13:45:16 mickeyl: sure Jun 13 13:45:19 koen: Ok, thanks. oprofile_cvs.bb will break in 5 to 24 hours then :) Jun 13 13:45:19 k Jun 13 13:45:20 Spyro: -f Jun 13 13:45:24 mickeyl: a patch would be better :) Jun 13 13:45:28 hah Jun 13 13:45:28 ;) Jun 13 13:45:31 Spyro: if its NULL ? Jun 13 13:45:45 koen: I'll see what i can do Jun 13 13:45:45 mallum: if whats null ? Jun 13 13:46:01 zecke: what about from the bitbake shell ? Jun 13 13:46:29 sirFred|tv, mickeyl, RP: W100:0 does not rotate Jun 13 13:46:40 it might sound like blasphemy but I only know the shell code Jun 13 13:47:19 Spyro: warning to stderr ? Jun 13 13:47:45 mallum: yeah Jun 13 13:47:56 mallum: although Im not sure it needs to be so vocal :) Jun 13 13:48:19 hrw|tv: That is known Jun 13 13:48:19 Spyro: no, there are some silly stderr stuff in tslib.c which I think is left over from debugging Jun 13 13:48:30 aha.. didnt know Jun 13 13:48:52 hrw|tv: It'll probably be tricky to fix as well. I need to finish working out what atilib is smoking... Jun 13 13:49:41 mickeyl: can I simply build (not fetch/unpack) something from the bitbake shell ? Jun 13 13:50:18 Spyro: unfortunately not yet. specifying of tasks and force is on the todo: / Jun 13 13:50:25 bugger Jun 13 13:50:30 indeed Jun 13 13:51:54 aah Jun 13 13:52:03 glib-2.0 doesn't depend on perl-native Jun 13 13:52:04 btw if I rebuild (say) tslib, do I need to rebuild task-gpe AND gpe-image or just gpe-image ? Jun 13 13:52:10 hence PERL_PATH gets picked up wrong Jun 13 13:55:38 Spyro: no Jun 13 13:55:52 mickeyl: no I do or no I dont ? Jun 13 13:55:54 you only need to rebuild when you change rdependencies Jun 13 13:56:09 i.e. when you change tslib to link against a new library Jun 13 13:56:25 mickeyl: what about gpe-image making me a rootfs ? Jun 13 13:56:58 rootfs'es are always rebuilt Jun 13 13:57:02 not overwritten Jun 13 13:57:15 mickeyl: what would prompt it to rebuild that? Jun 13 13:57:26 building task-gpe or gpe-image ? Jun 13 13:57:36 -image packages create rootfs'es Jun 13 13:57:40 ok Jun 13 13:57:41 task-gpe is just a bunch of packages Jun 13 13:58:23 whats task-gpe for ? Jun 13 13:58:47 just a virtual package for gpe related things Jun 13 13:58:57 what does that mean thoguh? Jun 13 13:59:07 mickeyl: python is sexy ;) Jun 13 13:59:08 ? Jun 13 13:59:09 (btw, wish me luck I have a new patched tslib/Xfbdev ;-) Jun 13 13:59:17 zecke: indeed it is .) Jun 13 13:59:21 mickeyl: whats a virtual package for gpe things ? Jun 13 13:59:24 hmm Jun 13 13:59:31 a virtual package is a package that just contains DEPENDS Jun 13 13:59:33 mickeyl: I will use my 'LIDL' train ticket to go to Karlsruhe Jun 13 13:59:35 no own content Jun 13 13:59:42 zecke: very good :) Jun 13 14:00:00 so why not use gpe-image to contain all those other gpe-packages ? Jun 13 14:00:46 Spyro: because this functional separation is good. it helps a) people to install gpe from scratch (by ipkg install task-gpe) and also help us to differentiate whether we tweak the image parameters or just the packages that belong to it Jun 13 14:01:02 mickeyl: any way to make a class 'internal'? prefix the name with '_' Jun 13 14:01:12 zecke: yah, that should work Jun 13 14:01:28 mickeyl: I see :) Jun 13 14:02:14 hmm Jun 13 14:02:35 why is perl missing do_stage() ? Jun 13 14:02:46 perl-native, that is Jun 13 14:02:46 Spyro: look into meta/meta-opie.bb - there are many *-task-* Jun 13 14:04:04 mickeyl: perl is to evil to be in staging Jun 13 14:04:18 koen: hmm... what then ? Jun 13 14:04:22 perl is needed by some packages Jun 13 14:04:45 those package are evil too Jun 13 14:04:51 heh Jun 13 14:04:53 death to interpreted languages! Jun 13 14:05:09 ok, let's remove glib-2 and everything that depends on it Jun 13 14:05:10 koen: nah, just perl Jun 13 14:05:11 anyway mono can be compiled just in time Jun 13 14:05:21 zecke: :) Jun 13 14:05:45 * koen hands zecke a frikandel speciaal Jun 13 14:06:01 yipee Jun 13 14:06:20 it's a bit sad that we need a host perl when we have a perl-native in OE Jun 13 14:07:00 perl-cross needs miniperl, at least that should get installed into staging Jun 13 14:07:11 heh, need to fix that bug where /usr/bin/autoconf, /usr/bin/automake, etc as built by oe have the hardcoded paths into staging if they were built after perl-native Jun 13 14:07:19 WTF? Jun 13 14:07:38 I rebuilt tslib and X and it still fails the same way Jun 13 14:07:39 cute bugs Jun 13 14:08:30 * mickeyl ln -s /usr/bin/perl5.8.5 /usr/bin/perl5 and forgets about the perl mess Jun 13 14:08:40 koen: I wanted to implement md5sum checks Jun 13 14:08:48 koen: but now I'm fixing a testcase Jun 13 14:09:32 * koen needs to install curry linux Jun 13 14:09:40 that has a md5 faking module Jun 13 14:09:52 with a hacked ftpd :) Jun 13 14:10:06 bah. it checked out a new CVS version. LOL Jun 13 14:14:45 wow. check this out... Jun 13 14:15:05 > I saw your website at http://www.mnementh.co.uk/ and was very impressed Jun 13 14:15:05 > with the work you have done on the toshiba e7xx platform Jun 13 14:15:05 > Jun 13 14:15:05 > We are a small company located in foobar and are in the process of Jun 13 14:15:05 > bringing a new video streaming technology to the internet Jun 13 14:15:26 wow Jun 13 14:15:28 they want a linux developer :-) Jun 13 14:15:34 I want a foobar postal address too! Jun 13 14:15:42 koen: hehe :) Jun 13 14:21:12 is foobar in Turkey? Jun 13 14:21:34 jacques: no, that's fubar Jun 13 14:22:44 ah lol I always get those confused :-P Jun 13 14:24:48 'night all Jun 13 14:25:35 03zecke123 * r251 10bitbake/lib/bb/__init__.py: (log message trimmed) Jun 13 14:25:35 BitBake: Jun 13 14:25:35 -Remove old code (python version >= 2.3 check) Jun 13 14:25:35 -Remove commented code (as we've a scm) Jun 13 14:25:35 -Fix most of the testcases (only one is failing) Jun 13 14:25:36 -by importing string Jun 13 14:25:38 -by removing new lines Jun 13 14:26:39 hrw|tv: Yes, there'are still a lot of work to do with w100. Jun 13 14:27:10 hrw|tv: The rotation issue is known ( by me ), I still had no time to look at it. Jun 13 14:27:41 sirFred|tv: I love you. Jun 13 14:27:53 ~lart linksys wrt54gs Jun 13 14:27:53 * ibot puts on some milking gloves. "All right, now, linksys wrt54gs, this won't hurt a bit...." Jun 13 14:28:11 hrw|tv: And about the gfx benchmark, yes, the circles are drawed really really slow. Jun 13 14:28:15 treke|home: :) Jun 13 14:29:09 hrw|tv: But look at the bright side: the driver is nearly able to handle a full opie session. Jun 13 14:29:25 sirFred|tv: nearly? Jun 13 14:29:48 hrw|tv: What problems has you found? Jun 13 14:29:58 (jackson was acquited for those that are interested) Jun 13 14:29:59 hrw|tv: Other than the bad gfx numbers and the rotation and the... Jun 13 14:30:19 sirFred|tv: just started opie and looked in ew places - not tested yet in normal life Jun 13 14:30:43 hrw|tv: ew places? Jun 13 14:30:52 few, probably Jun 13 14:30:52 few Jun 13 14:31:26 hrw|tv: Well, I'm working into it, but I have few time. Jun 13 14:31:34 got checkversion again... its good to have ^c available Jun 13 14:31:51 sirFred|tv: as we all.. 48h day would be great Jun 13 14:32:32 but only if worktime and sleeptime wouldn't increase Jun 13 14:32:47 yep Jun 13 14:33:19 hrw|tv: I also think that other people testing it, would be good for improving it. Jun 13 14:35:13 re Jun 13 14:35:56 hi CoreDump Jun 13 14:36:42 Looks like this company might financially support a port to the HP jornada... Jun 13 14:40:25 ok. flashing 2.6.12-rc5-mm1-r1 Jun 13 14:41:12 *bollocks*. Jun 13 14:41:24 fixing tslib seems not to have made everything better Jun 13 14:41:33 will look into it more later Jun 13 14:41:37 ttyl guys Jun 13 14:41:43 mickeyl: one question about updater.sh Jun 13 14:43:02 ya Jun 13 14:43:22 mickeyl: poodle one checks only for SL-5600 which is wrong as it can be B500 also Jun 13 14:43:34 c7x0 one check for c7x0, sl-5600, b500 Jun 13 14:43:46 so which one is for poodle? Jun 13 14:44:20 c7x0 should work for poodle, however i just have a ACK for the 5600 one Jun 13 14:44:40 so i recommend someone either testing the c7x0 one with poodle or adding the b500 to the 5600 one Jun 13 14:45:06 I will do second now due to lack of poodle Jun 13 14:45:11 heh Jun 13 14:45:11 thanks Jun 13 14:45:21 poodle owners are rare Jun 13 14:45:24 ~seen ced_ Jun 13 14:45:25 i haven't seen 'ced_', mickeyl Jun 13 14:45:28 ~seen ced Jun 13 14:45:29 ced <~ced@d83-177-204-95.cust.tele2.fr> was last seen on IRC in channel #debian, 80d 22h 9m 44s ago, saying: 'ok, so dont ask again ; )'. Jun 13 14:45:35 ~seen cedric Jun 13 14:45:36 cedric <~cedric@voulx.bluebugs.org> was last seen on IRC in channel #oe, 71d 11h 17m 44s ago, saying: 'did some one know why uclibc_cvs didn't use , why didn't you use SRC_URI += "svn://uclibc.org/trunk;module=uClibc ?'. Jun 13 14:45:45 71 days... Jun 13 14:45:46 cya tomorrow Jun 13 14:45:49 missing in action Jun 13 14:45:52 c'ya zecke Jun 13 14:46:23 cu zecke Jun 13 14:46:31 mickeyl: John Lenz has one... Jun 13 14:46:34 mia too Jun 13 14:47:15 ~poodle Jun 13 14:47:16 i heard poodle is sharp sl-b500/5600, or a dog Jun 13 14:49:02 worst zaurus models: a300 (nearly noone have it), poodle, corgi (for pxa250) Jun 13 14:49:14 CIA-4: 1.3583 pushed Jun 13 14:49:29 hmm Jun 13 14:49:37 ~lart zecke for reintroducing the deprecated string module in bb Jun 13 14:49:37 * ibot offers zecke some herring for reintroducing the deprecated string module in bb Jun 13 14:50:14 g'night folks Jun 13 14:50:30 n8 mickeyl Jun 13 14:50:31 night Jun 13 14:50:39 * hrw|tv goes sleep as well Jun 13 14:50:42 RP: tomorrow I'll add the patch from Todd to pxa-serialhack if you don't have any objections Jun 13 14:51:31 xmame 0.95 to add.. Jun 13 14:51:39 someone use xmame on arm anyway? Jun 13 15:05:35 umm pkg_postinst_ntpdate_nylon() Jun 13 15:05:47 any idea what "nylon" is supposed to be? machine? Jun 13 15:06:05 nylon is something related to openslug Jun 13 15:06:13 i see Jun 13 15:06:32 no it isnt Jun 13 15:06:54 well, this b0rks postinst for "our" ntpdate :) ie: no rcX.d symlinks Jun 13 15:07:26 iirc nylon is the meshcube distro Jun 13 15:07:48 #@DESCRIPTION: Distribution configuration for Nylon Mesh Linux distribution Jun 13 15:08:03 right. nothing to do with the nslu2 Jun 13 15:08:49 mickey|zzZZzz: No objections - looks like quite a sane patch Jun 13 15:08:58 mickey|zzZZzz: Although not mainline friendly :) Jun 13 15:10:25 "update-rc.d -s busybox-cron defaults" righto. What the hell is this doing in ntp's postinst heh. Jun 13 15:10:52 have a nice night Jun 13 15:10:59 n8 hrw|gone Jun 13 16:22:50 03CoreDump 07 * r1.3584 10openembedded/packages/ntp/ (ntp_4.1.2.bb files/ntp): ntp: Unb0rk postinst for ntpdate and add initscript for ntpd. Fixing Bug #85 Jun 13 16:50:21 'night all Jun 13 16:53:05 03RP 07 * r1.3573.1.1 10openembedded/packages/linux/ (10 files in 2 dirs): linux-oz-2.6: Upgrade to 2.6.12-rc6-mm1, add serial hack from Todd to enable correct numbering of pcmcia serial devices Jun 13 17:30:52 hey guys... Jun 13 17:31:08 do we have a workable Java VM in oe right now ? Jun 13 17:42:10 eww and yes i think so Jun 13 17:42:23 n8 Jun 13 17:42:31 night Jun 13 17:42:57 personally i wouldnt want to run it on an embedded device Jun 13 17:43:10 that was why they created java VM chips Jun 13 17:43:55 :) Jun 13 19:38:25 w100 patch has no rotation ;( Jun 13 19:39:26 oh there it is Jun 13 19:49:43 hmm. nope its using transformed now Jun 13 20:25:23 4/quit Jun 13 20:35:10 boo~! Jun 13 21:21:35 night Jun 13 22:31:09 hello all Jun 13 23:06:16 hi all **** ENDING LOGGING AT Mon Jun 13 23:59:56 2005