| 00:11.27 | *** join/#oe dkey| (~dkey@L0005P26.dipool.highway.telekom.at) |
| 00:21.54 | CIA-4 | 03cwiiis 07 * r1.2811.1.1 10openembedded/packages/ (2 files in 2 dirs): Merge |
| 00:21.54 | CIA-4 | 03cwiiis 07 * r1.2811.1.2 10openembedded/packages/ (4 files in 2 dirs): Bump PR on gpe-session-scripts (and a mvdir back and forth, sorry) |
| 00:21.55 | CIA-4 | 03cwiiis 07 * r1.2667.68.1 10openembedded/packages/meta/ (meta-gpe.bb meta-gpe-extras.bb): Complete move of gpe-task-games to meta-gpe. |
| 00:21.55 | CIA-4 | 03cwiiis 07 * r1.2667.68.2 10openembedded/packages/gpe-session-scripts/ (3 files in 2 dirs): Add collie key-mappings. |
| 00:26.47 | CIA-4 | 03pb 07 * r1.2814 10openembedded/packages/cairo/ (libsvg/expat.patch libsvg_cvs.bb): patch libsvg to use expat rather than libxml2 |
| 00:28.11 | CIA-4 | 03pb 07 * r1.2815 10openembedded/packages/meta/meta-gpe.bb: add gpe-soundbite to task-gpe-apps. remove useless miniclipboard. |
| 00:29.53 | CIA-4 | 03pb 07 * r1.2816 10openembedded/packages/meta/meta-gpe.bb: remove gweled from gpe-games one more time |
| 00:36.23 | Cwiiis | pb_: Ah, sorry for not noticing gweled in my merge, I'm guessing |
| 00:36.55 | pb_ | never mind |
| 00:37.37 | pb_ | I took it out this morning, because it drags librsvg and all the associated stuff into the image. |
| 00:38.26 | pb_ | If anyone would like to tweak it to use libsvg-cairo instead, it can probably go back in. |
| 00:39.32 | Cwiiis | pb_: Are the API's quite similar? |
| 00:40.13 | jacques | hi pb_ - I've been away for 11 days - do you recall if you committed the glibc-fp patch to openembedded ? |
| 00:40.38 | pb_ | Not as similar as all that, no. But, for simple uses, it shouldn't be that hard to convert the code. |
| 00:40.41 | andersee | pb_: oe can't seem to pull Tremor. I can see it via the xiph.org web interface, but it doesn't want to come down via svn. |
| 00:40.46 | pb_ | jacques: hm, no, I don't recall. let me have a look. |
| 00:40.51 | pb_ | andersee: that's very sad. |
| 00:42.05 | andersee | pb_: it appears there is a svn syntax error trying when oe tries to pull it |
| 00:42.26 | pb_ | how odd. what does it say? |
| 00:42.54 | bluelightning | I had the same thing the other day |
| 00:43.10 | pb_ | ah. what was the solution? |
| 00:43.15 | andersee | NOTE: package tremor-20041119-r1: task do_fetch started |
| 00:43.15 | andersee | NOTE: Fetch svn://svn.xiph.org/trunk;module=Tremor;rev=4573 |
| 00:43.15 | andersee | svn: Syntax error in revision argument 'svn' |
| 00:43.15 | andersee | NOTE: Task failed: Fetch failed: Tremor |
| 00:43.15 | andersee | ERROR: package tremor-20041119-r1: task do_fetch failed |
| 00:43.48 | bluelightning | pb_: I never found it |
| 00:44.15 | CIA-4 | 03pb 07 * r1.2817 10openembedded/packages/glibc/ (2 files in 2 dirs): add glibc-fp-byteorder patch for arm systems, courtesy Dwayne Fontenot |
| 00:44.21 | pb_ | jacques: there it is |
| 00:44.29 | bluelightning | pb_: I assumed bitbake was not compatible with the version of subversion I have installed... the .bb for tremor looked ok to me |
| 00:44.49 | jacques | pb_, whoo-hoo! thanks :-) |
| 00:45.10 | pb_ | andersee: as a workaround, you can probably get it from the CVS_TARBALL_STASH. |
| 00:45.37 | andersee | pb_: k. btw, could you check out http://www.codepoet.org/patch for me? I needed that to make opie-image build. |
| 00:45.51 | pb_ | fwiw, I get: |
| 00:45.59 | pb_ | NOTE: Fetch svn://svn.xiph.org/trunk;module=Tremor |
| 00:45.59 | pb_ | svn: Can't connect to host 'svn.xiph.org': Connection refused |
| 00:45.59 | pb_ | NOTE: Task failed: Fetch failed: Tremor |
| 00:46.31 | bluelightning | pb_: what version of subversion do you have installed? |
| 00:46.43 | pb_ | andersee: I'm not really au fait with opie-image. You should get mickey or kergoth to review that patch. |
| 00:46.55 | pb_ | svn, version 1.0.0 |
| 00:46.58 | andersee | $ svn --version | head -n1 |
| 00:46.58 | andersee | svn, version 1.0.9 (r11378) |
| 00:47.35 | andersee | pb_: k |
| 00:47.35 | bluelightning | I have 1.1.1 here... so I guess that's not it... ? |
| 00:47.37 | bluelightning | well actually it still could be |
| 00:48.45 | Cwiiis | jacques: What does that patch do? |
| 00:49.33 | jacques | Cwiiis, I found it in crosstool - it's absolutely required by armeb for fp to work right, and some builds of armel also require it |
| 00:49.52 | pb_ | bluelightning: what's the svn command line that bitbake uses for you? |
| 00:50.14 | jacques | for me it fixed tons of problems on the armeb nslu2 - everything from awk not working right to all my fp problems, to allowing me to build perl and get 100% test score |
| 00:50.38 | Cwiiis | jacques: armeb/armel? endianness I take it? |
| 00:50.50 | jacques | Cwiiis, yes big and little |
| 00:50.56 | Cwiiis | right |
| 00:51.02 | Cwiiis | I guess I should rebuild then :p hehe |
| 00:51.06 | Cwiiis | Well done on the patch :) |
| 00:52.00 | jacques | Cwiiis, the patch has extensive comments at the top IIRC - if you are on little-endian arm, you *may* not need it (tho it's not supposed to hurt any arm) |
| 00:54.31 | CoreDump|home | andersee: what's the error message w/o this patch? |
| 00:56.20 | andersee | CoreDump|home: opie-image fails because ${IMAGE_ROOTFS}/etc/ipkg.conf is missing |
| 00:57.08 | CoreDump|home | sounds like your ROM is missing ipkg-collateral |
| 00:57.19 | CoreDump|home | (which provides ipkg.conf) |
| 00:57.55 | andersee | CoreDump|home: ipkg-collateral was there, but wasn't installing anything |
| 00:58.39 | bluelightning | pb_: forgive me for being silly but how do I make bitbake show me the commands it is executing? BBDEBUG=1 doesn't do it |
| 00:58.53 | pb_ | bluelightning: bitbake -D |
| 00:59.00 | andersee | CoreDump|home: dunno why, but the patch carries on bravely with sane defaults should /etc/ipkg.conf go missing |
| 00:59.21 | CoreDump|home | andersee: i c. |
| 00:59.34 | CoreDump|home | for which machine are you building? |
| 00:59.54 | bluelightning | pb_: svn co -r svn 4573://svn.xiph.org/trunk/Tremor |
| 01:00.00 | bluelightning | no wonder it's not liking it |
| 01:00.03 | pb_ | hah |
| 01:00.32 | pb_ | DEBUG: Running svn co svn://svn.xiph.org/trunk/Tremor |
| 01:00.34 | CoreDump|home | nm, ipkg-collateral should produce ipkg.conf in any case |
| 01:00.35 | pb_ | that's what I get |
| 01:00.57 | andersee | CoreDump|home: openzaurus/tosa. I poked at ipkg-collateral but it didn't want to install /etc/ipkg.conf. |
| 01:00.58 | bluelightning | hmm... so yours isn't even specifying the rev |
| 01:01.02 | pb_ | yeah |
| 01:01.08 | pb_ | something odd does seem to be going on there. |
| 01:01.31 | pb_ | I guess the svn code in fetch.py is a bit broken. |
| 01:01.50 | andersee | CoreDump|home: Perhaps ipkg-collateral should blow chunks if ${IMAGE_ROOTFS}/etc/ipkg.conf is missing or 0 length |
| 01:02.09 | andersee | I'll look further at it |
| 01:02.22 | andersee | since I guess it is guilty |
| 01:02.29 | CoreDump|home | it is :) |
| 01:02.40 | CoreDump|home | whats your $MACHINE? |
| 01:03.05 | bluelightning | pb_: ah yes I see it, I'll make up a patch |
| 01:03.09 | andersee | CoreDump|home: MACHINE = "tosa" |
| 01:03.18 | pb_ | bluelightning: cool |
| 01:03.23 | andersee | CoreDump|home: Sharp Zaurus SL-6000L |
| 01:03.47 | reenoo_ | andersee: there's no ipkg.conf for tosa. probably want to add one |
| 01:03.55 | andersee | reenoo_: ahh |
| 01:04.01 | CoreDump|home | either I'm blind, or tosa doesn't have a "dest" file in collateral |
| 01:04.25 | andersee | I see |
| 01:04.41 | CoreDump|home | even so, a default (aka useless heh) ipkg.conf should be build |
| 01:05.13 | andersee | So, per my patch, is it reasonable to have openembedded/openembedded/packages/meta/opie-image.bb provide a sane default if /etc/ipkg.conf is missing? |
| 01:05.58 | andersee | or is that better done by ipkg-collateral? |
| 01:06.14 | pb_ | it should be done by ipkg-collateral. |
| 01:06.18 | CoreDump|home | indeed |
| 01:06.35 | pb_ | if that file is not showing up, either your ipkg-collateral.ipk is faulty, or something has gone wrong during rootfs construction. |
| 01:06.38 | CoreDump|home | or else every *-image file would need to cook their own ipkg.conf |
| 01:07.30 | pb_ | inspect it with dpkg-deb -c to find out which of those scenarios is in effect. |
| 01:08.08 | andersee | apparently though, ipkg-collateral needs explicit handling for every possible $MACHINE, or it leave ipkg.conf empty |
| 01:08.16 | pb_ | andersee: incorrect |
| 01:08.20 | CoreDump|home | andersee: not really empty |
| 01:08.30 | CoreDump|home | but "dest" entries are missing |
| 01:08.55 | pb_ | you won't have the "dest" entries for ram or cf, but you should be able to live without those. |
| 01:09.00 | andersee | CoreDump|home: at which point the egrep in packages/meta/opie-image.bb return an error. boom. |
| 01:09.10 | CoreDump|home | andersee: true |
| 01:09.22 | pb_ | corgi doesn't ship with those either, though I think that's just an oversight rather than a deliberate policy. |
| 01:09.46 | CoreDump|home | it probably should die because of that. On the other hand, ipkg.conf is essential |
| 01:10.02 | pb_ | ipkg.conf is essential, but those two "dest" entries are very far from essential. |
| 01:10.06 | bluelightning | pb_: http://pastebin.ca/3650 |
| 01:10.08 | CoreDump|home | s/should/shouldn't |
| 01:11.02 | pb_ | bluelightning: heh |
| 01:11.18 | CoreDump|home | andersee: that works as long as you never build a bootstrap image or gpe-image ;) |
| 01:11.23 | bluelightning | pb_: are you in a position to apply this or should I mail someone? |
| 01:11.28 | pb_ | bluelightning: I'll check it in |
| 01:11.33 | bluelightning | ok thanks |
| 01:12.05 | andersee | CoreDump|home: perhaps. Regardless, dying on an egrep is less than informative |
| 01:12.11 | CoreDump|home | but i agree that it should fail gracefully, but fail it should |
| 01:12.35 | CoreDump|home | maybe !test -e ipkg.conf && die "ipkg.conf is missing" or something like that |
| 01:13.20 | andersee | CoreDump|home: !test -s ${IMAGE_ROOTFS}/etc/ipkg.conf && die |
| 01:13.33 | CoreDump|home | right |
| 01:13.45 | andersee | ipkg-collateral is giving us a zero byte file |
| 01:14.03 | CoreDump|home | that shouldn't happen |
| 01:14.03 | pb_ | then you have a bad build of ipkg-collateral. |
| 01:15.00 | andersee | maybe that was just the 'mv's in opie-image.bb |
| 01:15.08 | andersee | probably that |
| 01:15.25 | andersee | started with nothing, then moved a 0 len file in place before bombing out |
| 01:15.43 | pb_ | <pb_> if that file is not showing up, either your ipkg-collateral.ipk is faulty, or something has gone wrong during rootfs construction. |
| 01:15.43 | RP | 'night all |
| 01:15.43 | pb_ | <pb_> inspect it with dpkg-deb -c to find out which of those scenarios is in effect. |
| 01:15.59 | pb_ | night rp |
| 01:16.07 | *** join/#oe offroadgeek (~offroadge@dsl027-191-243.sfo1.dsl.speakeasy.net) |
| 01:17.06 | CIA-4 | 03philb * r78 10bitbake/lib/bb/fetch.py: |
| 01:17.06 | CIA-4 | correct transposed protocol/revision parameters to svn checkout; |
| 01:17.06 | CIA-4 | patch courtesy Paul Eggleton |
| 01:17.51 | Cwiiis | neat, svn changes are reported here too? |
| 01:18.01 | CoreDump|home | andersee: just rebuild ipkg-collateral manually and see what happens |
| 01:18.13 | pb_ | Cwiiis: yes |
| 01:18.34 | Cwiiis | neat :) |
| 01:18.44 | andersee | CoreDump|home: dumb question. How do I force a package to rebuild? |
| 01:19.07 | CoreDump|home | bitbake -b $PATH_TO/.bb |
| 01:19.25 | CoreDump|home | first do a bitbake --cmd=clean -b $PATH_TO/.bb |
| 01:21.16 | andersee | k |
| 01:22.16 | CoreDump|home | it's best to make a shell function for that. |
| 01:22.48 | CoreDump|home | bbnew $PATH__TO_/.bb does the above for me |
| 01:27.13 | andersee | CoreDump|home: already did that |
| 01:27.14 | andersee | :-) |
| 01:27.18 | CoreDump|home | :) |
| 01:30.43 | *** join/#oe Goldarg (~goldarg@69-9-9-227.dakotacom.net) |
| 01:31.46 | Goldarg | Howdy folks, Is the GettingStarted doc in the wiki current? I'm attempting to follow it and I'm not able to setup bitbake |
| 01:32.16 | kergoth | yes, it's current. |
| 01:33.00 | Goldarg | When doing a './setup.py --prefix=/usr/local --installdata=/usr/local/share' I get a 'error: option --prefix not recognized' from setup.py in the bitbake dir |
| 01:33.24 | Goldarg | Debian testing, python 2.3.4 |
| 01:33.53 | kergoth | ah, right. make that ./setup.py install --prefix |
| 01:33.57 | kergoth | not ./setup.py --prefix |
| 01:34.20 | CIA-4 | 03pb 07 * r1.2818 10openembedded/packages/gpe-session-scripts/ (files/shepherd.xmodmap gpe-session-scripts_0.62.bb): tweak shepherd keymap in gpe-session-scripts |
| 01:38.52 | CIA-4 | 03CoreDump 07 * r1.2811.2.2 10openembedded/classes/rootfs_ipk.bbclass: rootfs_ipk: Check the logfile for error messages and error-out if necessary |
| 01:38.52 | CIA-4 | 03CoreDump 07 * r1.2811.2.1 10openembedded/packages/meta/opie-image.bb: opie-image: Fail gracefully if ipkg.conf is missing |
| 01:39.04 | Goldarg | Well that gets me something different at-least :) |
| 01:41.17 | Goldarg | Ahh good, All fixed with a little apt-get. |
| 01:44.22 | Goldarg | Makes me wish work was a debian shop some days. |
| 01:45.21 | CoreDump|home | i know your pain |
| 01:46.06 | jacques | pb_, still about? |
| 01:46.12 | pb_ | jacques: yo |
| 01:46.41 | Goldarg | As much as I love solaris and its brute ability to serve data debian makes life a lot easier for keeping systems current. |
| 01:47.24 | jacques | I notice you applied the glibc-fp-byteorder patch to all archs - I *thought* maybe I had seen something saying that might not be the best course. but then again now I can't find it :-\ |
| 01:47.42 | pb_ | jacques: it looked like it only touches files in sysdeps/arm, so it should be safe enough |
| 01:48.02 | jacques | yeah I just noticed that - thanks again :-) |
| 01:50.07 | andersee | heheh. mplayer-1.0pre6-r0 failed to build for arm. |
| 01:50.10 | andersee | | /usr/lib/gcc-lib/i486-linux/3.3.4/../../../crt1.o(.text+0x18): In function `_start': |
| 01:50.10 | andersee | | ../sysdeps/i386/elf/start.S:98: undefined reference to `main' |
| 01:50.10 | andersee | | collect2: ld returned 1 exit status |
| 01:50.27 | andersee | it seems to think i486 is useful for arm |
| 01:50.58 | reenoo_ | heh |
| 01:51.02 | reenoo_ | works fine here |
| 01:51.29 | reenoo_ | then again, I'm building against libsdl-x11 |
| 01:51.35 | andersee | I see |
| 01:51.40 | CoreDump|home | reenoo_: that's indeed the difference |
| 01:51.50 | CoreDump|home | it works w/ x11, not w/ qpe |
| 01:51.53 | andersee | it calls 'ccache gcc' for -o codec-cfg |
| 01:52.04 | CoreDump|home | andersee: pre2try2 works though |
| 01:57.10 | mickey|tv | pb_: how far did you get with your adding-usb-to-hostap? anything worth to submit to the hostap folks yet or did you give up prior to reaching that stage? |
| 01:57.24 | andersee | CoreDump|home: favor? |
| 01:57.48 | CoreDump|home | andersee: hmm? |
| 01:57.53 | pb_ | mickey|tv: not very far at all. I gave up on that and decided to pursue adding wireless-extensions support to wlan-ng instead. |
| 01:58.05 | mickey|tv | pb_: ah |
| 01:58.20 | mickey|tv | it looks like Pavel started with a prism_usb based on orinoco, but just a skeleton yet. |
| 01:58.26 | andersee | CoreDump|home: wanna do a 'mkdir -p openembedded/packages/ipkg/ipkg-collateral/tosa; cp openembedded/packages/ipkg/ipkg-collateral/collie/dest openembedded/packages/ipkg/ipkg-collateral/tosa' |
| 01:58.44 | pb_ | mickey|tv: ah right |
| 01:58.44 | andersee | and check that in |
| 01:58.46 | CoreDump|home | andersee: sure |
| 01:58.55 | andersee | CoreDump|home: thx |
| 02:01.00 | CIA-4 | 03CoreDump 07 * r1.2820 10openembedded/packages/ipkg/ipkg-collateral/tosa/dest: ipkg-collateral: Add tosa destinations |
| 02:01.13 | CoreDump|home | np |
| 02:03.47 | andersee | CoreDump|home: another dumb question. Based on CIA-4's msg you checked that in. But 'bk pull; /opt/bitkeeper/bk -r co -q' doesn't fetch anything for me. |
| 02:04.09 | CoreDump|home | ummm |
| 02:04.35 | pb_ | andersee: you're probably pulling from the anonymous tree, which lags the development sources by half an hour or so |
| 02:04.46 | andersee | pb_: ahhh |
| 02:04.57 | andersee | pb_: ok, that'd explain it |
| 02:04.58 | pb_ | either pull from oe-devel, or wait for the change to propagate to the anonymous one |
| 02:05.34 | andersee | no problem, just didnt know |
| 02:05.50 | pb_ | righto |
| 02:08.47 | CIA-4 | 03mickeyl 07 * r1.2808.1.1 10openembedded/packages/qte/ (4 files in 2 dirs): upgrade qte 2.3.9 snap to today and refresh the patches |
| 02:13.11 | reenoo_ | I'm off to bed. 'night all |
| 02:18.02 | *** part/#oe _narf_ (~noyb@fiftythree.hohertz.org) |
| 02:31.35 | *** join/#oe dalibor (~chatzilla@p5480D9C5.dip.t-dialin.net) |
| 02:33.34 | *** join/#oe W8TVI (~noel@209-124-42-78.dialup.torchlake.com) |
| 02:42.14 | *** join/#oe marcansoft (1337@80.26.222.195) |
| 02:43.27 | *** join/#oe BigAl (~bigal@dsl-203-113-232-253.ACT.netspace.net.au) |
| 02:50.29 | Cwiiis | Does gtk 2.6 build for anyone else? It fails to apply no-xwc.patch for me |
| 02:55.08 | *** part/#oe Goldarg (~goldarg@69-9-9-227.dakotacom.net) |
| 03:19.02 | *** join/#oe _kc8__ (~watcher@h-68-166-182-10.sfldmidn.covad.net) |
| 03:28.49 | *** join/#oe XEN (~XEN@d198-166-22-94.abhsia.telus.net) |
| 03:35.29 | *** join/#oe CoreDump|afk (~mhentges@xdsl-213-196-240-156.netcologne.de) |
| 03:36.48 | *** join/#oe CoreDump|home (~mhentges@xdsl-213-196-240-156.netcologne.de) |