| 00:00.02 | zecke | and even if I do PREFERRED_PROVIDER_virtual/... = "" it still complains |
| 00:00.08 | zecke | any hints where that comes from? |
| 00:00.17 | CIA-8 | 03RP 07 * r1.3217.1.1 10openembedded/packages/linux/ (2 files in 2 dirs): openzaurus-2.6: Add the devel work so far on tosa |
| 00:00.29 | ljp | hmm. except they jpg's arent jpgs |
| 00:01.30 | CIA-8 | 03RP 07 * r1.3224 10openembedded/packages/initscripts/initscripts_1.0.bb: initscripts: Finish off openzaurus-pxa-2.6 changes |
| 00:01.42 | mithro | i wonder how big the wayback machine is |
| 00:02.05 | ljp | nor are zip files |
| 00:09.38 | *** join/#oe Timelord0 (~TL@64.147.65.249) |
| 00:23.11 | *** join/#oe micropal_ (ident@139.80-202-82.nextgentel.com) |
| 00:28.46 | _keith | ls |
| 00:28.51 | _keith | oops |
| 00:29.10 | zecke | anyone here to help me with PREFERRED_PROVIDER? It must be something dumb?! |
| 00:33.26 | rpurdie | 'night all |
| 00:43.55 | _keith | Has anyone got Sound working on the SL-6000 at the moment? The syslog shows the initialization of "tosa AC97 Audio" but I see no sound related modules loaded or in /lib/modules for that matter. |
| 00:48.35 | *** join/#oe Speedy2 (~johnjacob@66-91-252-160.san.rr.com) |
| 01:44.50 | *** part/#oe Speedy2 (~johnjacob@66-91-252-160.san.rr.com) |
| 02:09.58 | *** join/#oe raster (~raster@vagw.valinux.co.jp) |
| 02:20.02 | *** join/#oe Timelord (~TL@64-147-65-174.access-4.intertex.net) |
| 03:02.59 | *** join/#oe treke|home (~ggilbert@69-175-32-221.ventca.adelphia.net) |
| 03:08.26 | *** join/#oe xep (misha@porter-kresge-apts-59-25.resnet.ucsc.edu) |
| 03:35.42 | *** join/#oe fishhead (~dropdead@c-24-3-15-166.client.comcast.net) |
| 03:35.48 | fishhead | I paste this because they are down to $13 at walmart |
| 03:35.57 | fishhead | http://elinux.sourceforge.net/wiki/JuiceBox |
| 03:50.46 | *** join/#oe rphillips (~rphillips@rphillips.developer.gentoo) |
| 04:03.38 | *** join/#oe treke|home (~ggilbert@69-175-32-221.ventca.adelphia.net) |
| 04:33.41 | *** join/#oe mithro (~tim@202.191.111.52) |
| 05:11.53 | *** join/#oe qip|Virus|da (~Thomas@pD95DB7DC.dip.t-dialin.net) |
| 05:19.46 | *** join/#oe SuN (unices@195-241-12-117-mx.xdsl.tiscali.nl) |
| 05:19.46 | *** join/#oe ljp (~lpotter@203.94.178.46) |
| 05:19.46 | *** join/#oe Timelord (~TL@64-147-65-174.access-4.intertex.net) |
| 05:25.37 | *** join/#oe Timelord (~TL@64-147-65-174.access-4.intertex.net) |
| 05:25.37 | *** join/#oe ljp (~lpotter@203.94.178.46) |
| 05:25.37 | *** join/#oe SuN (unices@195-241-12-117-mx.xdsl.tiscali.nl) [NETSPLIT VICTIM] |
| 06:04.46 | *** join/#oe caffinated (~dfgdf@2conn181.rupert.net) |
| 06:19.22 | *** join/#oe __law__ (~law@mail.stiftadmont.at) |
| 06:35.49 | *** join/#oe bluelightning (~bluelight@203-173-245-37.akl.ihugultra.co.nz) |
| 06:41.44 | *** join/#oe scanline_ (~micah@128.138.74.40) |
| 07:06.23 | *** join/#oe Luke-Jr (~luke-jr@207.192.221.172) |
| 07:16.24 | *** part/#oe caffinated (~dfgdf@2conn181.rupert.net) |
| 07:30.30 | *** join/#oe gints (~gints@159.148.213.135) |
| 07:31.39 | *** join/#oe treke|home (~ggilbert@69-175-32-221.ventca.adelphia.net) |
| 07:39.41 | cedric___ | ~logs |
| 07:39.41 | ibot | [logs] http://ibot.rikers.org/<channelname>/ where channelname is html encoded ie: %23debian, or stats for some channels at http://ibot.rikers.org/stats/<channelname>.html.gz for apt, ibot, jbot and purl. |
| 08:24.42 | *** join/#oe zedkatuf (~dfutak@82-32-58-45.cable.ubr08.azte.blueyonder.co.uk) |
| 08:32.41 | rpurdie | morning all |
| 08:33.02 | *** join/#oe Virusmaster (~Thomas@p5085CD5C.dip0.t-ipconnect.de) |
| 08:35.19 | koen | good morning all |
| 08:41.23 | cedric___ | morning |
| 08:42.28 | koen | hey cedric___ |
| 08:42.39 | RP__ | cedric___: update your copy of bitbake |
| 08:42.52 | RP__ | morning koen |
| 08:42.59 | koen | morning RP__ |
| 08:46.46 | hrw|work | morning everyone |
| 08:52.30 | koen | hey hrw|work |
| 08:53.52 | *** join/#oe schurig (~schurig@p5080A03F.dip0.t-ipconnect.de) |
| 08:54.59 | *** join/#oe micropal (ident@139.80-202-82.nextgentel.com) |
| 08:55.10 | cedric___ | RP__: what's the best way to create patches for the kernel, just make changes to the source and use diff on the original source? |
| 08:56.15 | schurig | cedric___: use the patcher tool |
| 08:56.20 | RP__ | cedric___: I cd into the kernel directory in the workdir and then use quilt on the files |
| 08:56.31 | schurig | cedric___: http://www.holgerschurig.de/patcher.html or "apt-get install patcher" |
| 08:56.59 | RP__ | Add a new patch and then keep your changes in that. Remove and save patch before wiping tmp! ;-) |
| 08:57.46 | RP__ | I used to use patcher but when oe switched to quilt, I also switched. They're pretty much the same thing... (I think you can tell oe to use patcher?) |
| 08:58.38 | schurig | RP__: use, "inherit patcher" in your local.conf file should do the trick |
| 08:59.04 | schurig | RP__: however, Patcher was originally designed for out-of-Bitbake-use, exactly for what was cedric___ asked: creating kernel patches |
| 08:59.29 | schurig | RP__: It doesn't matter if you use diff -urN or patcher or quilt to create patches, it's just that patcher is quite easy to use |
| 08:59.48 | RP__ | schurig: What's its advantage over quilt? They seem functionally equivalent to me... |
| 08:59.58 | schurig | RP__: for me, it's easier. |
| 09:00.26 | RP__ | Ok, just wondered. I like features of both :) |
| 09:00.38 | cedric___ | ok I'll have a look when I can get time from work... |
| 09:00.39 | schurig | RP__: also, when I made patcher, quilt was installing 30 or so different binaries into /usr/bin, and they where very hostily towards me when I wrote in the mailing list that this is not so good |
| 09:00.47 | schurig | So I wrote patcher |
| 09:00.57 | schurig | Now quilt is one-binary, too :-) |
| 09:01.33 | schurig | AFAIK quilt has more features, e.g. when it comes to merging. But for merely creating patches, patcher is easier. |
| 09:01.40 | schurig | But then again I'm biased :-) |
| 09:01.58 | RP__ | They're both very useful. Like I said, I like certain features of both |
| 09:05.42 | koen | so patcher is vi and quilt is emacs? |
| 09:05.48 | koen | ;) |
| 09:09.55 | *** join/#oe jbs (~jbs@213.58.72.17) |
| 09:10.10 | koen | hey Bernardo |
| 09:10.14 | schurig | koen: :-) |
| 09:10.22 | Bernardo | hi koen |
| 09:10.27 | Bernardo | hi guys |
| 09:16.23 | bluelightnin|tv | hi Bernardo |
| 09:25.08 | dave-osk | RP__: status update on the osk not talking via touchscreen ... it still doesn't |
| 09:25.33 | dave-osk | RP__: neither setting that environment variable (in any of several places) nor anything else affects it |
| 09:25.56 | dave-osk | RP__: I'm thinking somehow opie got built w/o touchscreen support! |
| 09:26.27 | RP__ | dave-osk: If you cat /dev/input/whatever and touch the touchscreen do you see anything? |
| 09:26.35 | koen | dave-osk: install tslib_tests and test with ts_print |
| 09:27.01 | RP__ | dave-osk: what koen said :) |
| 09:27.02 | dave-osk | RP__: o yes, "od -c /dev/input/X" works that way, for X != ts* |
| 09:27.35 | koen | most ts's in 2.6 use event* |
| 09:28.07 | dave-osk | RP__, koen: right, next time I build an image I'll include that. opie testing isn't the main effort. |
| 09:28.16 | RP__ | dave-osk: try ts_print from a console |
| 09:29.10 | dave-osk | RP__, koen: that's for tomorrow. the issue is in userland somewhere. 'night all! :) |
| 09:29.14 | RP__ | dave-osk: You don't have to rebuilt the image. oe will have built ts_tests as part of opie image - you just need to find the ipk and install it |
| 09:29.31 | *** join/#oe tigrux (~tigrux@styx.suse.cz) |
| 09:30.03 | RP__ | dave-osk: There will be a file in /tmp/deploy/ipk called tslib-tests, copy it to the Z and do ipkg install file.ipk... |
| 09:41.42 | *** join/#oe hrw (szczepan@blue.stone.pl) |
| 09:41.59 | hrw | re |
| 09:45.47 | Bernardo | hi hrw|work |
| 09:46.35 | hrw|work | Bernardo: kdepimpi 2.0.7 have recurrent todos |
| 09:50.19 | Bernardo | I just saw that on the forums |
| 09:50.27 | Bernardo | but he broke the fonts, right? |
| 09:52.13 | *** join/#oe mickey|bbl (~mickey@141.2.14.1) |
| 09:53.25 | hrw|work | fix is easy but I refuse to do it for him |
| 09:53.54 | Bernardo | You and your bad temper... :) |
| 09:53.55 | hrw|work | Bernardo: unpack kdepimpi, replace microkde/ofontselector.cpp with libopie2/libcore/ one and build |
| 09:54.12 | RP__ | morning mickey|bbl |
| 09:54.35 | hrw|work | it is like zautrix said: "libopie has nothing usefull for me" - but he copied atleast fontreq from it |
| 09:54.56 | Bernardo | I'll have to remember that when I get home. Just got to install 2.0.6 this morning, my wife almost killed me we where leaving so late |
| 09:54.56 | *** join/#oe Timelord (~TL@64-147-65-174.access-4.intertex.net) |
| 09:55.28 | cedric___ | mickey|bbl: how's yr tosa 2.6 kernel? |
| 09:58.40 | RP__ | cedric___: Dirk Opfer sent us some patches which move tosa forward an awful lot |
| 09:58.59 | treke|home | I still don't get why people go to the effort of patching out libopie |
| 09:59.12 | RP__ | I checked them into oe yesterday - untested... |
| 09:59.16 | treke|home | it's less work to just install the library |
| 09:59.37 | hrw|work | treke|home: because its fun to give user 2MiB in size app with 5MiB libraries |
| 10:00.17 | cedric___ | what's the tosa graphic chip? |
| 10:01.31 | RP__ | cedric___: tc6393 maybe? |
| 10:01.57 | cedric___ | ok, was just wondering if it was the same than the poodle |
| 10:03.21 | hrw|work | cedric___: collie has one, poodle another, c7x0 has ati w100, tosa has toshiba, sl-3000 has pxa27x controler |
| 10:04.56 | treke|home | what a mess |
| 10:05.15 | RP__ | Just what I thought :) |
| 10:05.26 | treke|home | err. wrong channel |
| 10:05.31 | treke|home | but you can think that anyways |
| 10:05.52 | RP__ | it made sense anyway. It is a mess :) |
| 10:05.56 | treke|home | hehe |
| 10:06.36 | treke|home | I should go to sleep |
| 10:11.02 | koen | 'morning mickeyl |
| 10:11.10 | mickeyl | morning folks |
| 10:11.36 | mickeyl | say good bye to your known machine configurations heh |
| 10:11.39 | mickeyl | no more husky |
| 10:11.42 | mickeyl | no more shepherd |
| 10:11.44 | mickeyl | no more corgi |
| 10:12.42 | Bernardo | too bad no more collie... |
| 10:13.01 | mickeyl | no no |
| 10:13.04 | mickeyl | collie is cool |
| 10:13.06 | mickeyl | collie will stay |
| 10:13.13 | mickeyl | i just don't like the other dogs |
| 10:13.13 | mickeyl | heh |
| 10:13.40 | hrw|work | mickeyl: you dont like shepherd, tosa too? |
| 10:13.44 | RP__ | mickeyl: One problem I considered - how do we differentiate between the different flash sizes on those devices? |
| 10:14.02 | Bernardo | and simpad? 2.6. is still far away for that one? |
| 10:14.18 | Bernardo | which I assume is what you're talking about... |
| 10:15.04 | RP__ | mickeyl: I did think SUBMACHINE= would probably be the way to solve that if it was a problem... |
| 10:15.48 | hrw|work | RP__: its 64 on other then husky and 128 on husky? |
| 10:15.52 | mickeyl | RP: it is no problem |
| 10:16.09 | mickeyl | RP: atm. it only makes a difference if the FLASH is less than 24 or larger |
| 10:16.22 | *** join/#oe lardman (~lardman@138.38.73.178) |
| 10:16.24 | mickeyl | for now i just set FLASH_MEMORY_SIZE to 32 which is the common demonitator |
| 10:16.26 | RP__ | mickeyl: Ok. Just wondered :) |
| 10:16.29 | mickeyl | ibot: spell denominator |
| 10:16.30 | ibot | 'denominator' may be spelled correctly |
| 10:17.38 | mickeyl | ERROR: |
| 10:17.39 | mickeyl | * |
| 10:17.39 | mickeyl | * Sorry, There is no dedicated configuration for the Sharp SL-C750, |
| 10:17.39 | mickeyl | * because of the similarities of the clamshell series. Use MACHINE = "c7x0". |
| 10:17.39 | mickeyl | * |
| 10:17.43 | mickeyl | here we are |
| 10:18.30 | hrw|work | Not updating idcache due to dups. |
| 10:18.30 | hrw|work | Check FAILED, contact BitMover. |
| 10:18.30 | hrw|work | resolve: RESYNC directory left intact. |
| 10:18.31 | hrw|work | argh |
| 10:21.17 | lardman | How can I pass an env var to be included when Makeconf.in is turned into Makeconf? |
| 10:21.36 | lardman | I do something similar for make: oe_runmake 'R_EXE_NATIVE = ${STAGING_BINDIR}/R' |
| 10:21.52 | hrw|work | lardman: EXTRA_OEMAKE? |
| 10:21.53 | lardman | but don't know what turns makeconf.in into makeconf |
| 10:22.11 | hrw|work | lardman: you can also export variable |
| 10:22.20 | lardman | hrw|work: won't that pass it to make? |
| 10:22.35 | hrw|work | lardman: yep - I misunderstood |
| 10:22.40 | hrw|work | lardman: export variable maybe? |
| 10:22.49 | hrw|work | or own do_compile |
| 10:22.52 | CIA-8 | 03mickeyl 07 * r1.3225 10openembedded/ (40 files in 16 dirs): consolidate collie, shepherd, husky into the common machine c7x0. kernel 2.6 handles all three machines, for kernel 2.4 we build three different kernels for every -image build. |
| 10:23.02 | lardman | I need to replace a variable in the makeconf.in file so it's passed down the chain |
| 10:23.23 | lardman | hrw|work: Can I export the var from inside the .bb file? |
| 10:23.58 | hrw|work | lardman: I saw it in few bb files |
| 10:24.50 | hrw|work | lardman: are you registered in openzaurus.org website? |
| 10:24.57 | lardman | hrw|work: ok, I'll try a grep, I can't use do_compile as the var is hard-coded by that point (it's hard-coded in the configure script) |
| 10:25.03 | lardman | hrw|work: I think so |
| 10:26.02 | koen | btw, can anyone with OZ-site access correct the install instructions? |
| 10:26.21 | koen | teh sl-c700 does *not* have a reset button like it says in the instructions |
| 10:26.25 | hrw|work | lardman: I'm thinking about post to oz-users,oz-devel ML to notify users that without them website will be like it is now |
| 10:26.37 | hrw|work | koen: send me corrected version so I will change it |
| 10:26.45 | koen | ok |
| 10:27.27 | lardman | I brought that up a long while back, but I don;'t think I had the required clout to be believed |
| 10:28.09 | hrw|work | lardman: I saw that someone setup wiki with some kind of oz manual even |
| 10:28.11 | lardman | I assume you're takling about the fact that 5600/c750/c700 are all listed with the same method? |
| 10:28.11 | koen | yes |
| 10:28.11 | koen | the *wrong* method |
| 10:28.12 | lardman | Excuse my poor typing, I've broken my wrist |
| 10:28.32 | mickeyl | cedric___: remind me, SD on poodle is via Locomo or via PXA ? |
| 10:29.17 | lardman | hrw|work: on oz.org, or elsewhere? |
| 10:29.35 | hrw|work | other place |
| 10:29.45 | hrw|work | lardman: and I dont have link here ;( |
| 10:29.55 | lardman | hrw|work: there was a push on the zug to get things documented, but as oz.org went live then we sort of waited to see what would happen about wiki access |
| 10:31.57 | cedric___ | mickeyl: via PXA, I think, I've got the detection at the moment, it power on twice and power off strait after |
| 10:32.15 | mickeyl | cedric___: aha. very good. then collie will probably be the only model we won't support w/ SD |
| 10:32.44 | hrw|work | mickeyl: or we get kergoth to write them |
| 10:32.51 | *** join/#oe zinga (~arnaudb@186.80-203-227.nextgentel.com) |
| 10:32.52 | mickeyl | heh |
| 10:32.52 | mickeyl | yeah |
| 10:32.55 | mickeyl | but that's not very likely |
| 10:33.10 | hrw|work | mickeyl: he is only one who can do it anyway |
| 10:33.35 | cedric___ | mickeyl: I'm trying to build a bootstrap image with all the modification on the machine thingy, so after that I'll try to fix the SD and test the pcmcia |
| 10:40.03 | hrw|work | Consistency check passed, resolve complete. |
| 10:40.08 | hrw|work | finally... |
| 10:40.39 | hrw|work | bitkeeper sometimes behave strange... but "rm somewhere/SCCS/s.something" can help when good used |
| 10:44.40 | mickeyl | *cough* |
| 10:44.43 | mickeyl | that's pretty hackish |
| 10:44.45 | mickeyl | :) |
| 10:46.24 | hrw|work | mickeyl: but it works quite good when bk say "contact bitmover" just because some dups |
| 11:02.57 | mickeyl | hmm libsdl 1.2.8 |
| 11:03.43 | hrw|work | noone cooperated with upstream to get qpe part inside? |
| 11:04.32 | mickeyl | no one |
| 11:04.39 | mickeyl | quite unfortunate |
| 11:05.53 | hrw|work | I mailed with subapplet author - there will be two versions - "lite + config" and "full" |
| 11:06.22 | hrw|work | my work on its sources ;) |
| 11:07.11 | mickeyl | nice |
| 11:07.53 | mickeyl | although I'd rather have someone working on making the taskbar hosting multiple rows of applets |
| 11:08.06 | mickeyl | or implementing a windows-like notification bar |
| 11:08.20 | mickeyl | where inactive applets are hidden |
| 11:08.20 | hrw|work | opie launcher code is .... |
| 11:08.22 | hrw|work | argh |
| 11:08.26 | mickeyl | true enough |
| 11:08.40 | hrw|work | creating anycol was enough for me |
| 11:12.02 | hrw|work | how was your work on opie2 summary? |
| 11:12.16 | mickeyl | no time for that |
| 11:12.41 | mickeyl | we probably just put up what we have this week |
| 11:12.46 | mickeyl | and let it evolve |
| 11:14.35 | hrw|work | OT - how to speedup launching konq under kde 3.3? |
| 11:14.58 | mickeyl | <-- galeon user |
| 11:15.45 | hrw|work | I was galeon user in 1.2 times. |
| 11:16.04 | hrw|work | 1.3 was going simplier and simplier... so I switched in firefox |
| 11:22.07 | *** join/#oe zecke (~ich@83.169.146.2) |
| 11:31.16 | *** join/#oe qfh (~qfh@ppp-62-245-163-25.mnet-online.de) |
| 11:38.34 | *** join/#oe reenoo (~rwagner@x20.informatik.uni-bremen.de) |
| 11:38.57 | reenoo | morning |
| 11:39.36 | mickeyl | moin reenoo |
| 11:41.08 | hrw|work | hi Rene |
| 11:42.58 | reenoo | moin mickeyl |
| 11:43.03 | reenoo | hey Marcin |
| 11:52.04 | *** join/#oe hrw (szczepan@blue.stone.pl) |
| 11:56.38 | *** join/#oe hrw (szczepan@blue.stone.pl) |
| 11:57.48 | hrw | re |
| 12:04.33 | *** join/#oe TyggerBob (~myname@64.26.144.98) |
| 12:23.02 | hrw | koen: fixed installation |
| 12:24.05 | koen | hrw: thanks! |
| 12:24.28 | koen | I felt very stupid when I couldn't find the reset button :) |
| 12:24.32 | hrw | ;) |
| 12:24.59 | *** join/#oe ade|desk (~adavey@194.200.143.249) |
| 12:25.07 | ade|desk | afternoon folks |
| 12:26.14 | koen | hey ade|desk |
| 12:44.40 | CIA-8 | 03mickeyl 07 * r1.3226 10openembedded/packages/libsdl/libsdl-qpe_1.2.7.bb: libsdl-qpe: add dependency to libopie2 |
| 12:52.03 | *** join/#oe HCl_ (hcl@titania.student.utwente.nl) |
| 12:52.28 | CIA-8 | 03mickeyl 07 * r1.3227 10openembedded/packages/linux/linux-openzaurus-2.6.11-rc4/ (defconfig-tosa defconfig-c7x0): oz-kernel-2.6: consolidate c7x0 |
| 13:02.54 | CIA-8 | 03mickeyl 07 * r1.3228 10openembedded/packages/qpf-fonts/qpf-hunky_0.3.0.bb: add qpf-fonts hunkysans and hunkyserif. patch courtesy Gints Polis |
| 13:03.48 | *** join/#oe ade|des1 (~adavey@194.200.143.249) |
| 13:13.31 | *** join/#oe katossi (~Sara@dsl-082-083-061-050.arcor-ip.net) |
| 13:14.07 | katossi | anyone has recently compiled atk under oe?? |
| 13:14.12 | koen | katossi: the quick solution is to symlink it, but I'd like to know why it searches in the wrong staging dir |
| 13:14.38 | katossi | so do you get the same errors? |
| 13:15.11 | *** join/#oe [g2] (~g2@rdu57-8-035.nc.rr.com) |
| 13:15.11 | zecke|sick | I need someone to look at my oe setup |
| 13:15.23 | zecke|sick | PREFERRED provider refuses to work and I'm too blind to find it |
| 13:16.34 | *** part/#oe katossi (~Sara@dsl-082-083-061-050.arcor-ip.net) |
| 13:16.45 | *** join/#oe katossi (~Sara@dsl-082-083-061-050.arcor-ip.net) |
| 13:18.07 | mickey|lunch | does it appear in bitbake -c showdata ? |
| 13:21.51 | zecke|sick | mickeyl: it says nothing to do?! |
| 13:22.11 | zecke|sick | wait |
| 13:24.00 | katossi | how do I change an oe package to be compiled with debugging enabled? |
| 13:24.40 | mickeyl | read local.conf |
| 13:24.43 | mickeyl | .sample |
| 13:26.09 | koen | is there a smart way to detect TARGET_FPU? |
| 13:26.15 | mickeyl | detect ? |
| 13:26.19 | mickeyl | in how far |
| 13:26.29 | mickeyl | oh you mean on a complete image |
| 13:26.35 | koen | no, in a .bb |
| 13:26.43 | mickeyl | hmm |
| 13:26.46 | mickeyl | shell |
| 13:26.54 | koen | sharp-aticore.bb only work if ${TARGET_FPU}=soft |
| 13:27.51 | koen | with anything other the fetch fails |
| 13:27.58 | mickeyl | you would need an anonymous function |
| 13:28.01 | mickeyl | raising a skip package exception or so |
| 13:28.38 | koen | SRC_URI = "http://www.mithis.com/zaurus/AtiCore.${TARGET_FPU}.tar.bz2" |
| 13:28.46 | koen | that's the problem |
| 13:29.03 | koen | AtiCore.tar.bz2 and AtiCore.soft.tar.bz2 exist |
| 13:29.21 | koen | but AtiCore..tar.bz2 obviously doesn't exist |
| 13:29.52 | mickeyl | i see |
| 13:30.00 | CIA-8 | 03mickeyl 07 * r1.3229 10openembedded/conf/distro/familiar-0.8.1.conf: fix typo in familiar 0.8.1.conf |
| 13:30.07 | mickeyl | mithro better fixes that |
| 13:31.06 | koen | ah, thanks for fixing that typo |
| 13:31.12 | katossi | mickeyl are you going to fosdem 2005? |
| 13:31.24 | mickeyl | if i managed to book a flight and a hotel this week, then yes |
| 13:31.29 | mickeyl | s/managed/manage/ |
| 13:31.42 | mickeyl | i'm scheduled as a speaker so i better appear :) |
| 13:31.51 | koen | indeed |
| 13:32.36 | koen | finally, one report done, two to go |
| 13:32.49 | mickeyl | bbl, meeting |
| 13:33.02 | katossi | myckeyl cool, I hope i can go too |
| 13:33.35 | katossi | I just read your name in fosdem's developers rooms |
| 13:34.18 | katossi | mickey switching debugging on should be an easy task then, isn't it? |
| 13:34.29 | katossi | it's not I don't want to take a look at local.conf |
| 13:34.43 | koen | mostly setting INHIBIT_PACAKGE_STRIP (or similar) |
| 13:34.47 | hrw | katossi: in conof/local.conf you have to uncomment 2 options |
| 13:34.47 | katossi | but right now I'm in my girlfriend's ibook, with no linux at all |
| 13:34.54 | hrw | katossi: comments say all |
| 13:35.00 | koen | me on powerbook/OSX |
| 13:35.07 | katossi | really? |
| 13:35.53 | katossi | unfortunately if I put any oe stuff here, she will kill me |
| 13:36.03 | katossi | it's very picky with his "sweety" |
| 13:36.04 | koen | OE runs om my linux workstation |
| 13:36.28 | katossi | so you don't do any development on it, do you? |
| 13:36.59 | hrw | gints: ping |
| 13:36.59 | koen | not really |
| 13:37.00 | katossi | I found that java developing in mac os x is great |
| 13:37.10 | koen | (as I don't really do development at all) |
| 13:37.38 | katossi | koen do you just do packing, or what? |
| 13:37.56 | koen | mostly |
| 13:38.15 | katossi | because if you do any debug, debugging is developing, in some ways... |
| 13:38.41 | hrw | I do packaging, fixes in packages, sometimes code fixes to get something working |
| 13:44.44 | gints | hrw: pong |
| 13:45.14 | hrw | gints: which method did you use to generate husky qpf fonts? |
| 13:45.29 | hrw | gints: I tried few times but my Z does not want to see them |
| 13:45.31 | gints | used opie for that... |
| 13:46.18 | gints | hrw: just to be sure you have updated fontdir... |
| 13:46.26 | hrw | gints: updated |
| 13:46.55 | hrw | anyway I'm building qte 2.3.10 for my desktop. maybe will try to build makeqpf for zaurus |
| 13:47.34 | gints | oe just got bb file for hunky... |
| 13:47.50 | gints | and for me it works fine... |
| 13:48.39 | gints | I also can give my built ipk.. if you want so... |
| 13:48.46 | hrw | yep I know - thats why I'm asking you |
| 13:48.53 | hrw | I already built that ipks |
| 13:48.59 | gints | ah.. |
| 13:50.38 | hrw | usually I use verdana ttf/9 by opie-freetype |
| 13:50.56 | gints | I used this link for creating fonts... http://opie.handhelds.org/cgi-bin/moin.cgi/CreateFontsForOpie |
| 13:50.57 | hrw | on qvga it is best font for me |
| 13:51.04 | hrw | thx |
| 13:51.34 | gints | np |
| 13:51.55 | hrw | Connecting to moria.ionkov.net[68.36.169.182]:80... failed: Connection refused. |
| 13:51.57 | hrw | ;( |
| 13:52.09 | hrw | could you put makeqpf-arm somewhere? |
| 13:53.17 | gints | sory I have it in home... |
| 13:53.25 | gints | just evening... |
| 13:53.32 | hrw | spox |
| 13:54.06 | *** join/#oe Tygger_Bob (~myname@CPE00045a265f64-CM000f9f7c58fa.cpe.net.cable.rogers.com) |
| 13:55.03 | Tygger_Bob | hehehe... |
| 13:59.53 | *** join/#oe pb_ (~pb@192.208.47.123) |
| 14:01.06 | *** join/#oe stevenh (~xev@65.167.23.2) |
| 14:04.42 | CIA-8 | 03mickeyl 07 * r1.3230 10openembedded/ (7 files in 2 dirs): add necessary .bb files to build all c7x0 kernels in one run |
| 14:06.27 | hrw | "include ../linux/openzaurus-pxa_2.4.18-rmk7-pxa3-embedix20030509.bb" - "include openzaurus-pxa_2.4.18-rmk7-pxa3-embedix20030509.bb" wont be easier? |
| 14:16.53 | *** join/#oe woglinde (woglinde@pD9E7F56A.dip.t-dialin.net) |
| 14:45.37 | *** join/#oe sgm (~sgm@mevissen.demon.nl) |
| 14:46.04 | *** join/#oe hcl (hcl@titania.student.utwente.nl) |
| 14:47.46 | *** join/#oe hcl (hcl@titania.student.utwente.nl) |
| 14:57.39 | *** join/#oe drw- (~drw@c-67-172-218-120.client.comcast.net) |
| 15:02.28 | *** join/#oe Tygger-Bob (~myname@64.26.144.98) |
| 15:03.27 | chouimat | morning |
| 15:06.06 | *** join/#oe molivier (~mac@f06v-9-178.d1.club-internet.fr) |
| 15:08.45 | *** join/#oe prpplague (~dave@mailhost.amltd.com) |
| 15:13.14 | *** join/#oe TyggerBob (~myname@CPE00045a265f64-CM000f9f7c58fa.cpe.net.cable.rogers.com) |
| 15:17.35 | pb_ | hey prpplague |
| 15:21.06 | prpplague | pb_: hey hey |
| 15:21.10 | prpplague | pb_: how goes it? |
| 15:21.24 | prpplague | pb_: have a good pub crawl this past weekend? hehe |
| 15:21.45 | pb_ | prpplague: not bad, not bad. I'm in sunny MA at the moment. |
| 15:26.13 | sgm | Who can help me with building a native (i686) version of gpe-image? |
| 15:26.54 | CosmicPenguin | sgm: why don't you try it, and let us know if it breaks? |
| 15:26.58 | chouimat | http://www.askmen.com/jokes/current/index.html |
| 15:27.13 | CosmicPenguin | damn |
| 15:27.17 | sgm | CosmicPenguin: well, it brakes. that's why I'm asking |
| 15:27.26 | CosmicPenguin | I was going to bring my tuxscreen in and make a shannon target in OE |
| 15:27.40 | pb_ | sgm: best to tell us what actually goes wrong. |
| 15:28.03 | sgm | pb_: after some troubles, I'm stuck with xserver-kbuild now |
| 15:28.18 | sgm | I had the problem that xentensions was to old for it |
| 15:29.10 | sgm | I solved that by setting the DEPENDS of xentension to xextensions-${PV} |
| 15:29.19 | sgm | I hope that is the correct way to do it. |
| 15:29.43 | sgm | this still has the problem that Xext does use xextentions 1.0.something |
| 15:29.52 | sgm | But that is not the issue now |
| 15:30.08 | sgm | Now the build fails because KdInitTouchScreen |
| 15:30.41 | sgm | symbol cannot be found in xserver-kdrive-0.0cvs20050214-r8/xserver/hw/kdrive/epson/epson13806stub.c |
| 15:30.59 | sgm | that symbol is only needed when TOUCHSCREEN is defined... |
| 15:31.13 | pb_ | disable the epson driver in EXTRA_OECONF |
| 15:31.27 | pb_ | or patch epson13806stub.c to check #ifdef TOUCHSCREEN |
| 15:31.52 | sgm | pb_: that file does already contain that #ifdef |
| 15:32.31 | hrw | hm.. I need tslib static |
| 15:32.32 | pb_ | in that case, you need to find out why the symbol doesn't exist |
| 15:32.58 | sgm | pb_: because tslib isn't used natively I guess |
| 15:33.14 | pb_ | KdInitTouchScreen is not part of tslib |
| 15:33.51 | sgm | Where is it then? |
| 15:34.15 | pb_ | I don't know. You have the source and grep, though, so you should be able to find it. |
| 15:35.02 | sgm | pb_: maybe as a last resort. But TOUCHSCREEN should not be defined any, I guess? |
| 15:35.24 | pb_ | Check your EXTRA_OECONF. |
| 15:36.10 | sgm | xserver-kdrive_cvs.bb:EXTRA_OECONF = "--enable-composite --disable-xinerama" |
| 15:36.12 | sgm | hmmm |
| 15:37.06 | sgm | and it's configure is also quite strange in detecting whether or not to use TOUCHSCREEN |
| 15:37.30 | *** join/#oe Tygger_Bob (~myname@64.26.144.98) |
| 15:37.30 | koen | xinerama can be handy in combination with the CF xga card |
| 15:38.13 | zecke | wow that works? |
| 15:38.13 | koen | judging from my apache logs, my uni is starting to use OE as well |
| 15:38.32 | koen | zecke: they say there is an X-server for that card |
| 15:38.55 | sgm | koen: I would be happy to be able to play with gpe on an ordinary i686... |
| 15:39.12 | koen | but since I don't have E250 burning in my wallet I can't try it :) |
| 15:39.25 | pb_ | sgm: if you think the configure script is wrong, that would be an issue for #freedesktop |
| 15:39.34 | hrw | hmm.. I never built libs by hand so have a question - how to build tslib static? (I want to build makeqpf static and all what I miss is tslib static now) |
| 15:39.55 | koen | sgm: if you want x86 you can just use the makefile in gpe CVS |
| 15:40.08 | sgm | pb_: Well, I miss the knowledge to really find that out. Maybe it is the setup that is created by OE |
| 15:40.57 | CIA-8 | 03RP 07 * r1.3231 10openembedded/packages/linux/linux-openzaurus_2.6.11-rc4.bb: openzaurus-2.6: Updates for c7x0 change and fix for tosa |
| 15:41.47 | sgm | koen: I would like to use OE for i686 and ARM, using the same SW. So it might take longer, but then we can use it all. |
| 15:42.21 | koen | I see |
| 15:43.10 | sgm | koen: it is just that it drives me nuts that I'm already working since last week sunday (not yesterday) to get it working |
| 15:43.18 | sgm | and that even with good support from here. |
| 15:43.23 | koen | you might want to include multimachine.inc, otherwise you'll need two trees |
| 15:43.58 | pb_ | he won't be able to share anything worthwhile between x86 and arm anyway, so two separate build trees would be best |
| 15:43.59 | sgm | koen: when the i686 works, then I'll try that. |
| 15:44.25 | koen | pb_: true, dind't think of that |
| 15:44.43 | koen | I also use it as an extra form of instrumentation |
| 15:45.29 | koen | it also exposes hardcoded paths from time to time |
| 15:57.32 | CIA-8 | 03mickeyl 07 * r1.3230.1.1 10openembedded/conf/machine/zaurus-clamshell-2.4.conf: fix stupid typo in zaurus-clamshell-2.4 conf |
| 16:00.04 | RP__ | mickey|meeting: I wondered what I'd done there for a minute! :) |
| 16:02.35 | *** join/#oe Tygger-Bob (~myname@CPE00045a265f64-CM000f9f7c58fa.cpe.net.cable.rogers.com) |
| 16:03.02 | *** join/#oe HuntsMan (~hunts@pc-148-118-120-200.cm.vtr.net) |
| 16:06.18 | treke|home | <PROTECTED> |
| 16:06.41 | pb_ | crumbs |
| 16:06.43 | pb_ | how does that work? |
| 16:06.51 | pb_ | (repartitioning, not larting) |
| 16:07.19 | treke|home | not entirely sure, but there is a menu in the updater.sh for his distro |
| 16:07.41 | pb_ | funky |
| 16:07.59 | cedric___ | it'll be nice to have a single partition like on their distro |
| 16:08.29 | treke|home | seems like a loss of a useful setup to me |
| 16:09.11 | treke|home | with two partitions, I don't need to worry about data loss during a reflash. |
| 16:09.25 | pb_ | right |
| 16:09.39 | pb_ | and, undesirable though it might be, most of our users wind up reflashing quite often |
| 16:10.56 | RP__ | I suspect you can repartition just by flashing a larger nand image and telling the kernel the different size |
| 16:11.33 | pb_ | How do you tell the kernel, though? |
| 16:12.01 | treke|home | I'd be willing to bet it includes a nasty hack |
| 16:12.19 | RP__ | pb_: Through the command line nand partition params |
| 16:12.32 | pb_ | RP__: right, but where does that come from? |
| 16:12.51 | RP__ | pb_: They get put into the kernel when you compile it... |
| 16:12.59 | *** join/#oe tux_mike (~mike@zelda.tuxnami.org) |
| 16:13.14 | pb_ | ah, so you need to flash a custom kernel for each different partition setup? |
| 16:13.16 | treke|home | couldn't they just write over the spot in the kernel where it is stored with the new params? |
| 16:13.39 | RP__ | pb_: that's one option. I'll bet you can find and edit them in a kernel image as well |
| 16:13.48 | pb_ | treke|home: in theory, but the kernel is compressed so patching it in place would be tricky |
| 16:14.27 | RP__ | pb_: Not all of the kernel is compressed. The params might not be... |
| 16:14.48 | pb_ | RP__: if this is CONFIG_CMDLINE, it traditionally lives in the compressed region. |
| 16:15.25 | pb_ | I guess you could hack it to live in the non-compressed area somehow. |
| 16:15.25 | RP__ | pb_: fair enough. You need a different kernel per setup then... |
| 16:15.27 | pb_ | but it'd probably be easier to just unzip the kernel, patch it, then re-zip it. |
| 16:15.36 | RP__ | or perhaps you can edit the ones the sharp bootloader passes... |
| 16:16.01 | treke|home | I'll ask him before he's drug out into the street and shot |
| 16:19.51 | mickey|meeting | they rewrite the PARAMS section |
| 16:20.01 | mickey|meeting | hence altering the angel bootloader args |
| 16:20.06 | mickey|meeting | for zimage2 |
| 16:20.22 | mickey|meeting | and also for their zimage which obeys the bootloader params |
| 16:20.24 | pb_ | aha |
| 16:20.36 | mickey|meeting | pretty ugly thing to do |
| 16:20.38 | mickey|meeting | imo |
| 16:21.01 | cedric___ | pb_/ RP__: if u want to keep some data why not using media card for that, so I think a big partition is nice and u get a better compression of the jffs2 |
| 16:21.29 | pb_ | cedric___: I don't want to keep "some" data. I want to keep _all_ my data. |
| 16:22.09 | CIA-8 | 03mickeyl 07 * r1.3233 10openembedded/packages/linux/ (4 files): make openzaurus-pxa a .inc file to indicate that it isn't a direct build target |
| 16:22.11 | pb_ | I don't understand your comment about compression. |
| 16:22.23 | cedric___ | ok then store al lyr data under your card |
| 16:23.11 | pb_ | That would be unacceptable too, because then the machine would be useless unless that one card was fitted. |
| 16:23.16 | cedric___ | if u got a bigger flash then I "think" the partition as jff2 type will more compressed |
| 16:23.42 | pb_ | What leads you to think that? |
| 16:26.07 | cedric___ | maybe I'm wrong but just by the factor of a bigger size, the compression algorithum will be more efficient... but that's just a thought... |
| 16:26.34 | cedric___ | never try it so... |
| 16:27.05 | cedric___ | a big chance I'm talking sh*t |
| 16:27.55 | pb_ | afaik, jffs2 compression is all intra-block. |
| 16:28.58 | cedric___ | which means? |
| 16:29.56 | pb_ | each block is compressed without reference to others. |
| 16:30.15 | cedric___ | ok |
| 16:36.30 | cedric___ | but it'll be nice to create a such utility if people want to resize their partition (base on their personal preferences) |
| 16:45.20 | *** join/#oe Timelord0 (~TL@64-147-65-70.access-3.intertex.net) |
| 16:51.23 | zecke|bath | koen: how does familiar 0.8.1 build for you? |
| 16:53.00 | koen | zecke|bath: apart from the atk thing pretty good |
| 16:53.47 | CIA-8 | 03rjt 07 * r1.3165.1.4 10openembedded/packages/gpe-session-scripts/gpe-session-scripts_0.62.bb: oops. forgot to up PR for the change to packages/gpe-session-scripts/files/xdefaults.patch |
| 16:53.50 | CIA-8 | 03rjt 07 * r1.3165.1.3 10openembedded/packages/sudo/ (3 files in 2 dirs): make sudo install setuid |
| 16:53.53 | CIA-8 | 03rjt 07 * r1.3165.1.2 10openembedded/packages/python/python-pygtk-0_0.6.12.bb: fix rebuilding of pygtk |
| 16:53.56 | CIA-8 | 03rjt 07 * r1.3165.1.1 10openembedded/packages/gpe-session-scripts/files/xdefaults.patch: changes to keep font size sane with new rxvt-unicode |
| 16:54.21 | zecke|bath | I'm struggling two days to build a toolchain |
| 16:54.52 | zecke|bath | what is the preferred glibc version for familiar? |
| 16:55.09 | pb_ | 2.3.2+cvs20040726 |
| 16:55.27 | pb_ | which, coincidentally, is also the default in oe |
| 16:55.35 | zecke|bath | | checking installed Linux kernel header files... TOO OLD! |
| 16:55.35 | zecke|bath | | configure: error: GNU libc requires kernel header files from |
| 16:55.46 | pb_ | zecke|bath: sounds like your kernel header files are... TOO OLD |
| 16:55.55 | *** join/#oe jamie (~jamie@softmodem.org) |
| 16:56.00 | zecke|bath | when building glibc-intermediate-2.3.2+cvs20040726-r14 |
| 16:56.04 | *** join/#oe marcansoft (1337@80-29-18-137.adsl.nuria.telefonica-data.net) |
| 16:56.07 | zecke|bath | pb_: is it really referring to the host headers? |
| 16:56.15 | pb_ | no, it is referring to the target headers. |
| 16:56.17 | pb_ | are you on x86? |
| 16:56.32 | zecke|bath | yes |
| 16:56.35 | pb_ | x86 target, that is |
| 16:56.42 | ade|des1 | why does glibc need to refer to kernel headers? |
| 16:56.45 | zecke|bath | Linux 2.0.10 or later |
| 16:56.56 | pb_ | ade|des1: er, because the kernel headers define the system call interface |
| 16:56.58 | zecke|bath | I really think that debian unstable ships newer headers? |
| 16:57.10 | pb_ | zecke|bath: it does not use the headers from your host system |
| 16:57.22 | pb_ | it uses the target headers within your oe build directory |
| 16:58.03 | zecke|bath | ok so a depend is missing somewhere (I've no *header* package build) |
| 16:58.10 | *** join/#oe Tygger_Bob (~myname@64.26.144.98) |
| 16:58.39 | pb_ | zecke|bath: ah. it should have built linux-kernel-headers.oe, or some such |
| 16:58.43 | anomaly_ | lol |
| 16:58.51 | zecke | ah that was the name |
| 16:58.59 | pb_ | linux-libc-headers, actually |
| 16:59.15 | pb_ | linux-libc-headers_2.6.8.1.bb |
| 16:59.15 | pb_ | that one |
| 17:04.02 | *** join/#oe rphillips (~rphillips@rphillips.developer.gentoo) |
| 17:04.53 | *** join/#oe TyggerBob (~myname@CPE00045a265f64-CM000f9f7c58fa.cpe.net.cable.rogers.com) |
| 17:18.26 | *** join/#oe tomimo (~kurre@a84-231-39-238.elisa-laajakaista.fi) |
| 17:18.35 | *** join/#oe micropal_ (ident@139.80-202-82.nextgentel.com) |
| 17:24.28 | lardman | Can somone tell me which procedure (like gnu-configize, perhaps) generates the makeconf file from makeconf.in please? |
| 17:30.10 | mickeyl | lardman: never heard of such a file. looks non standard to me. |
| 17:30.26 | lardman | which one? makeconf.in? |
| 17:30.30 | mickeyl | you should build the package on x86 to get an idea of how the build system works |
| 17:30.32 | mickeyl | yes, makeconf.in |
| 17:30.58 | pb_ | generally, foo is generated from foo.in by the configure script |
| 17:31.00 | CIA-8 | 03zecke 07 * r1.3231 10openembedded/packages/glibc/ (2 files): |
| 17:31.01 | CIA-8 | glibc 2.3.2+cvs2004726: |
| 17:31.01 | CIA-8 | <PROTECTED> |
| 17:31.01 | CIA-8 | <PROTECTED> |
| 17:31.08 | lardman | oh, I thought makconf.in -> makeconf; makconf + configure.in -> configure |
| 17:31.37 | pb_ | zecke: ah, very good |
| 17:32.08 | mickeyl | pb_: http://handhelds.org/scap/port.27903.png |
| 17:32.15 | mickeyl | looks like someone is using scap |
| 17:32.22 | mickeyl | but it's corrupted as well |
| 17:32.36 | koen | 130.89....? |
| 17:32.49 | mickeyl | yes |
| 17:32.51 | pb_ | it seems |
| 17:32.52 | koen | that would be me :) |
| 17:32.56 | pb_ | I wonder what version of scap they have |
| 17:32.58 | pb_ | koen: ah |
| 17:34.53 | *** join/#oe zwi (~zwi@216.88.131.43) |
| 17:35.08 | *** join/#oe mallum (~mallum@192.208.47.35) |
| 17:35.44 | zecke | hmm bk repair created a cset of a change I didn't intent to be comitted |
| 17:36.30 | koen | btw, scap also segfaults on my corgi |
| 17:37.17 | pb_ | yeah, that's a kernel bug |
| 17:37.27 | pb_ | what version of scap do you have? |
| 17:38.02 | koen | scap - 0.1-r0 - |
| 17:39.35 | lardman | Mickeyl: Ah, it looks like configure uses the file; does gun_configize run the configure script? |
| 17:39.44 | mickeyl | no |
| 17:39.47 | mickeyl | oe_runconf runs the scripts |
| 17:39.52 | mickeyl | gnu-configize just prepares it |
| 17:40.01 | lardman | from configure.in? |
| 17:40.15 | lardman | by running autotools, etc |
| 17:40.16 | mickeyl | no. in general it is recommended to just let it regenerate using standard autotools build calss |
| 17:40.33 | mickeyl | check autotools.bbclass for some details |
| 17:40.58 | lardman | ok, that's fine, I can pass the env var I need to oe_runconf, that will hopefully allow R to package correctly |
| 17:41.07 | pb_ | lardman: all that gnu-configize does is install config.sub and config.guess in the source directory |
| 17:41.21 | lardman | pb_: okay, thanks |
| 17:42.44 | CIA-8 | 03zecke 07 * r1.3233 10openembedded/packages/nonworking/kde/arts_1.2.2.bb: |
| 17:42.44 | CIA-8 | ARTS: |
| 17:42.44 | CIA-8 | <PROTECTED> |
| 17:45.37 | *** join/#oe SirFred (~mteira@85.Red-81-34-170.pooles.rima-tde.net) |
| 17:45.52 | SirFred | Hi. |
| 17:46.48 | koen | hey SirFred |
| 17:46.55 | SirFred | koen: Hi. |
| 17:47.01 | koen | pb_: is that scap up to date? |
| 17:49.21 | pb_ | as far as I know, yeah |
| 17:49.24 | pb_ | does it call fbset? |
| 17:52.37 | koen | yes |
| 17:52.58 | zecke | Somebody wants to test the icecream bbclass? |
| 17:53.00 | pb_ | then it's up to date |
| 17:53.09 | pb_ | check what arguments it's sending to the server. |
| 17:54.12 | *** part/#oe ade|desk (~adavey@194.200.143.249) |
| 17:58.44 | *** join/#oe guillermo (~guillermo@dsl-082-083-061-050.arcor-ip.net) |
| 18:05.06 | koen | pb_: the awk fails, forcing RES to "640x480" yields http://handhelds.org/scap/port.9353.png |
| 18:05.40 | pb_ | why does the awk fail? |
| 18:06.45 | koen | don't know, but running the script gives and empty RES, doing fbset | awk "/geometry/ { print \$2 "x" \$3 }" gives me 640480 |
| 18:07.36 | pb_ | without the "x"? that's pretty weird |
| 18:07.44 | pb_ | sounds like you need to do some debugging |
| 18:07.53 | koen | does awk handle all thos """" correct? |
| 18:08.11 | pb_ | dunno |
| 18:08.20 | pb_ | does it help if you change the outside ones to ''? |
| 18:08.39 | koen | root@corgi:~# fbset | awk '/geometry/ { print \$2 "x" \$3 }' |
| 18:08.40 | koen | awk: cmd. line:1: Unexpected token |
| 18:09.06 | pb_ | you might need to remove those \ as well |
| 18:09.24 | pb_ | the shell won't do $-substitution inside `` |
| 18:09.40 | koen | root@corgi:~# fbset | awk '/geometry/ { print $2 "x" $3 }' |
| 18:09.40 | koen | 640x480 |
| 18:09.51 | pb_ | very good |
| 18:10.52 | koen | testing the complete script |
| 18:12.03 | koen | we have a winner: http://handhelds.org/scap/port.9811.png |
| 18:12.30 | pb_ | cool |
| 18:12.34 | koen | this scap script works: http://pastebin.ca/5813 |
| 18:14.33 | *** join/#oe marcansoft (1337@80-29-18-191.adsl.nuria.telefonica-data.net) |
| 18:16.05 | pb_ | great. will you check that in? |
| 18:16.25 | *** join/#oe zwi (~zwi@69.25.212.5) |
| 18:17.22 | CIA-8 | 03mickeyl 07 * r1.3229.3.1 10openembedded/packages/linux/ (2 files): reintroduce openzaurus-pxa.bb - tosa and poodle still need it |
| 18:21.51 | *** join/#oe treke (~ggilbert@tinman.treke.net) |
| 18:25.29 | *** join/#oe dougt (~dougt@adsl-63-198-216-19.dsl.snfc21.pacbell.net) |
| 18:38.46 | *** join/#oe J_ (~john@host-83-146-11-165.bulldogdsl.com) |
| 18:47.48 | dougt | where is CVSDATE usually defined for a package? |
| 18:48.13 | *** join/#oe florian (~fuchs@p5092955F.dip0.t-ipconnect.de) |
| 18:48.28 | florian | re |
| 18:50.44 | *** join/#oe J_ (~john@host-83-146-11-165.bulldogdsl.com) |
| 18:51.16 | zecke | dougt: that depends |
| 18:51.29 | treke | usually defined in the distro.conf or local.conf |
| 18:51.32 | zecke | dougt: if it is not 'your' package in conf/local.conf |
| 18:52.03 | dougt | so, local.conf alway overrides the distro one... good. ;-) |
| 18:52.28 | zecke | dougt: yes and no, position depends |
| 18:52.39 | dougt | sure. |
| 18:58.21 | *** join/#oe reenoo_ (~r@pD9528CB7.dip.t-dialin.net) |
| 18:59.44 | reenoo_ | evening |
| 19:01.35 | mickeyl | evening |
| 19:01.55 | tux_mike | afternoon |
| 19:04.44 | reenoo_ | 'nabend mickeyl |
| 19:04.49 | reenoo_ | hi tux_mike |
| 19:05.30 | tux_mike | hey |
| 19:12.47 | *** join/#oe gremlin[it] (~gremlin@194-185-220-78.f4.ngi.it) |
| 19:12.53 | gremlin[it] | good evening :) |
| 19:16.06 | *** join/#oe chouimat (~dieu@r2351064.cidc.net) |
| 19:18.35 | *** join/#oe Shdwdrgn (Picard@sourpuss.net) |
| 19:18.55 | gremlin[it] | got a problem with atk |
| 19:19.08 | koen | symlink it |
| 19:19.10 | mickeyl | we too. some of kergoths variable changes broke it |
| 19:19.45 | gremlin[it] | :( |
| 19:19.58 | pb_ | gremlin[it]: bk pull |
| 19:20.17 | gremlin[it] | already did |
| 19:20.42 | pb_ | then you'll have to tell us what error you're experiencing. |
| 19:20.45 | pb_ | atk builds fine for me. |
| 19:20.51 | gremlin[it] | ok |
| 19:21.03 | gremlin[it] | >>>gettext source directory '/mnt/hda/oe/stuff/build/tmp/staging/share/glib-2.0/gettext' doesn't exist<<< |
| 19:22.20 | pb_ | what version of glib-2.0-native do you have? |
| 19:22.22 | gremlin[it] | days ago i got problem with xrender, to build i set CVSDATE_xserver-kdrive=20050207, resolved ? |
| 19:23.04 | gremlin[it] | pb_ i'm tring to build gpe-image familiar-0.8.1 for ipaq h3600 |
| 19:23.24 | CIA-8 | 03RP 07 * r1.3237 10openembedded/packages/linux/linux-openzaurus_2.6.11-rc4.bb: openzaurus-2.6: fix corgi reference |
| 19:23.37 | pb_ | gremlin[it]: that's very interesting. what version of glib-2.0-native do you have? |
| 19:24.05 | gremlin[it] | sorry i suppose that could be an answer ;) ... wait |
| 19:25.12 | gremlin[it] | pb_ tmp/work/glib-2.0-native-2.6.2-r0 ... better ? |
| 19:26.05 | pb_ | your tree is out of date. |
| 19:26.07 | pb_ | bk pull. |
| 19:27.21 | gremlin[it] | how far bk repository was updated ??? i did lenn than 30min ago ... and now there is a glib patch :) |
| 19:28.39 | CosmicPenguin | gremlin[it]: the bk repo is almost always changing, especially during the hours when the primary developers are awake |
| 19:29.28 | gremlin[it] | i know ... but really i got the error just before login :) ... |
| 19:30.57 | gints|home | and you see how fast problems are solved here :) |
| 19:31.21 | gremlin[it] | living on the edge :) :) :) |
| 19:31.54 | *** join/#oe Skyhusker_ (~fiam@80.103.43.90) |
| 19:32.56 | pb_ | zecke: btw, please don't bump PR on glibc for things like your DEPENDS change. |
| 19:33.31 | reenoo_ | heh |
| 19:33.47 | zecke | pb_: argh |
| 19:34.28 | zecke | pb_: rule of thumb, bump PR on change of packaging? |
| 19:35.01 | koen | stuff that affects the outcome of the build |
| 19:35.06 | pb_ | zecke: the rule of thumb is that you bump PR whenever you make a change that will affect the output .ipk packages. |
| 19:35.19 | pb_ | or the files in staging, or any other visible output |
| 19:35.56 | pb_ | adding DEPENDS doesn't generally do that, so there is no point in forcing everyone to recompile the thing and install it on their target systems. |
| 19:36.32 | pb_ | for small packages it doesn't matter too much, but rebuilding and reinstalling glibc causes a significant annoyance for a lot of people |
| 19:38.09 | reenoo_ | what about RDEPENDS vs. RRECOMMENDS changes and bumping PR btw? it doesn't really influence the output, just the control file... |
| 19:38.40 | mickeyl | the control file is output |
| 19:38.47 | pb_ | RDEPENDS and RRECOMMENDS both go into the ipkg control file, so they do affect the output. |
| 19:39.04 | pb_ | even DEPENDS _can_ affect the output in some cases, but generally it doesn't. |
| 19:39.11 | reenoo_ | hmm, right. |
| 19:51.06 | CIA-8 | 03reenoo 07 * r1.3238 10openembedded/packages/bridge-utils/bridge-utils_1.0.4.bb: bridge-utils: bump PR due to RDEPENDS->RRECOMMENDS change. |
| 19:58.46 | *** join/#oe bipolar (~bipolar@146.145.26.90) |
| 19:59.02 | *** join/#oe SirFred (~mteira@85.Red-81-34-170.pooles.rima-tde.net) |
| 20:02.56 | koen | mickeyl: the initscript .bb isn't multimachine friendly due to the use of STAGING_KERNEL_DIR, do you now a better solution than inheriting module-base? |
| 20:04.08 | pb_ | koen: what's your handhelds.org bugzilla account name? |
| 20:05.10 | koen | koen@handhelds.org |
| 20:07.39 | pb_ | thanks |
| 20:08.33 | mickeyl | koen: not offhand. I'll think about it |
| 20:08.52 | koen | thanks |
| 20:11.10 | CIA-8 | 03reenoo 07 * r1.3241 10openembedded/packages/tinc/tinc_1.0.2.bb: tinc: demote kernel-module-tun to RRECOMMENDS |
| 20:11.14 | CIA-8 | 03reenoo 07 * r1.3240 10openembedded/packages/chillispot/chillispot_0.98.bb: chillispot: demote kernel-module-tun to RRECOMMENDS |
| 20:11.18 | CIA-8 | 03reenoo 07 * r1.3239 10openembedded/packages/htb-init/htb-init_0.8.5.bb: htb-init: demote kernel-module-* to RRECOMMENDS |
| 20:12.15 | reenoo_ | there are a couple more left. I'll take care of them tomorrow |
| 20:12.24 | pb_ | cool |
| 20:12.53 | *** join/#oe TyggerBob (~myname@CPE00045a265f64-CM000f9f7c58fa.cpe.net.cable.rogers.com) |
| 20:13.15 | koen | mickeyl: and using MACHINE=c7x0 multimachine.inc breaks since /data/build/oe/build/tmp-all-OZ/staging/c7x0-linux/kernel/kernel-source doesn't exist |
| 20:13.31 | *** join/#oe sgm (~sgm@mevissen.demon.nl) |
| 20:13.49 | *** join/#oe darkschneider (~gab@81-208-36-80.fastres.net) |
| 20:13.54 | *** join/#oe zap (~zap@217.170.93.9) |
| 20:14.20 | koen | I haven't thought of a clean solution for that |
| 20:17.54 | pb_ | eh, how are you ending up with c7x0-linux? |
| 20:18.42 | pb_ | oh, right, multimachine.inc sticks PACKAGE_ARCH in there. |
| 20:20.49 | koen | indeed |
| 20:21.39 | koen | inheriting module-base 'fixes' most of the issues, but MACHINE is not really MACHINE here |
| 20:31.25 | *** join/#oe segv (~manihp@zux221-079-166.adsl.green.ch) |
| 20:34.30 | *** join/#oe florian (~fuchs@p5092955F.dip0.t-ipconnect.de) |
| 20:46.22 | *** join/#oe dkey (~dkey@193.170.62.146) |
| 20:55.31 | *** join/#oe jamie (~jamie@modem-1605.porcupine.dialup.pol.co.uk) |
| 21:11.41 | *** join/#oe _keith (~keith@CPE-61-9-204-1.nsw.bigpond.net.au) |
| 21:30.41 | dave-osk | koen: re that OSK touchscreen issue ... ts_print is quite happy once I set the environment variable |
| 21:31.06 | dave-osk | koen: ... but opie itself doesn't see it. where's that supposed to get set? |
| 21:31.11 | *** join/#oe Pigi (~NoOne@host3-70.pool21345.interbusiness.it) |
| 21:31.25 | dave-osk | koen: ... and is there a good fix for the situation where "event1" may be wrong, because |
| 21:31.33 | koen | /etc/profile.s/tslib.sh |
| 21:31.45 | dave-osk | koen: ... for example the usb keyboard and mouse got enumerated first? |
| 21:32.19 | dave-osk | koen: tslib.sh is what I tried second; it didn't seem to have any effect. I'll double check. |
| 21:32.48 | koen | dave-osk: I heard something about ioctls, but you'll have to ask the tslib gurus |
| 21:32.59 | pb_ | detect-stylus --device |
| 21:33.05 | pb_ | see the h3900 tslib.sh |
| 21:33.36 | koen | ah yea, forgot about that one |
| 21:34.35 | dave-osk | koen: yes, it was set in tslib.sh and it's not working. |
| 21:34.39 | dave-osk | TSLIB_TSDEVICE=/dev/input/event1 |
| 21:34.39 | dave-osk | export TSLIB_TSDEVICE |
| 21:34.42 | dave-osk | yes? |
| 21:34.55 | dave-osk | ... looking for h3900 tslib.h |
| 21:34.59 | pb_ | and you sourced that file? |
| 21:35.03 | pb_ | (or /etc/profile) |
| 21:36.33 | dave-osk | yes it was in /etc/profile.d where the opie startup script seems to suck it in |
| 21:36.50 | dave-osk | found the h3900 one, looks like it should be the default on all platforms ... ;) |
| 21:37.36 | RP__ | dave-osk: Not all systems use the input system. I'd agree it should be default though |
| 21:38.40 | dave-osk | RP__: the problem I seem to have with most OE packages is they expect an iPAQ, and maybe LK 2.4 ... |
| 21:39.11 | dave-osk | RP__: some don't, of course. the pcmcia startup script chokes on non-Intel CPUs for example; |
| 21:39.15 | *** join/#oe beewoolie (~beewoolie@florence.buici.com) |
| 21:39.50 | dave-osk | RP__: doesn't know what to do with an ARM926EJS ! assumes CPU controls what pcmcia controller to use. nope. |
| 21:40.29 | dave-osk | ... ok, I'm off to try that new tslib trick. thanks for the help, all! |
| 21:41.05 | RP__ | dave-osk: Ipaq/2.4 linux/openzaurus is where oe has come from... Packages shouldn't be too dependent on that. If they are and it can be changed there would be no objection to fixing them... |
| 21:41.34 | pb_ | dave-osk: feel free to send patches to teach pcmcia-cs about your particular PCIC. |
| 21:42.01 | pb_ | with a 2.6 kernel, I guess hotplug should be handling all that stuff anyway. |
| 21:42.21 | dave-osk | RP__: yes, I realize that. but it takes time to sort through all the undocumented OE stuff to find how to fixit! |
| 21:42.27 | pb_ | so, maybe we just need to wrap the strongarm/xscale stuff in a version check so it only fires on 2.4. |
| 21:43.17 | dave-osk | pb_: in this case I suspect that OMAP-specific things in hotplug or udev could help. |
| 21:44.13 | pb_ | dave-osk: right |
| 21:44.18 | dave-osk | pb_: there's also all the hotplug stuf for ipaq sleeves; specific to ipaq, not strongarm or xscale (nslu2 sleeves? :) |
| 21:44.26 | RP__ | dave-osk: I'm not an oe expert and have the same problem. Documentation is a problem :-/ |
| 21:44.51 | pb_ | dave-osk: sure, but that stuff is harmless. on non-ipaq, you will never get any sleeve events generated, and sleeve.agent will just sit quietly doing nothing. |
| 21:45.18 | pb_ | the flip side to that is that the ipaq doesn't, for example, have pci. |
| 21:45.25 | dave-osk | RP: when is doc not a problem for developers? :) just with OE it's moreso. |
| 21:46.00 | dave-osk | pb: right, but when I want a rootfs without all sorts of irrelevant clutter, OE needs to be chastized. |
| 21:46.11 | reenoo_ | dave-osk: it becomes less of a problem as soon as you have enough developers |
| 21:46.18 | RP__ | dave-osk: I'm not disagreeing :) |
| 21:46.35 | dave-osk | pb: like, oh, bluetooth ... not on this board, why is it the default for everyone?? |
| 21:46.49 | pb_ | dave-osk: in general, it's easier for maintainability to have one hotplug package that runs on all machines, rather than a million different variants with some files left out. |
| 21:47.08 | dave-osk | reenoo: agreed. more eyes. more contributors. more modes of usage stretching things. |
| 21:47.57 | dave-osk | pb: sort of. clutter isn't free. but let me tell you, I'll never have S390 DASD devices on this hardware! |
| 21:48.15 | dave-osk | what I'd like to see is some more meta-description ... |
| 21:48.23 | pb_ | dave-osk: if you desperately need to have sleeve and bluetooth support removed for your platform, you can do that, but I'd encourage you not to do so unless there is a burning reason. |
| 21:48.37 | dave-osk | saying "this board has hardware X, Y, Z, ..." |
| 21:49.16 | dave-osk | so that OE can say "when X, include <these packages>"; "when Y, include <those others>" ... |
| 21:49.41 | dave-osk | pb: nothing now is burning ... but it's all examples of |
| 21:50.10 | dave-osk | pb: assumptions **throughout the metadata** that don't match my target(s) |
| 21:50.34 | pb_ | alternatively, if you wanted to split hotplug up into multiple packages that could be selectively installed, that would be fine too. |
| 21:50.39 | dave-osk | pb: OE would be better if it offered config options there |
| 21:51.18 | dave-osk | pb: thing is, the sleeve additions are OE-specific; and there are zaurus/2.4 analogues |
| 21:51.19 | RP__ | dave-osk: To be fair, oe does produce a fairly minimal filesystem and does go to great lengths to do so... |
| 21:51.20 | pb_ | dave-osk: patches are welcome |
| 21:51.44 | dave-osk | pb: the standard hotplug doesn't add those! |
| 21:52.01 | pb_ | right, but one day it will. |
| 21:52.04 | treke | no it doesn't, we add them in because the distribution we are shipping uses them |
| 21:52.50 | dave-osk | treke: I have no problem with ipaq-specific features in ipaq distros ... |
| 21:53.03 | dave-osk | treke: ... or zaurus-specific features in zaurus distros |
| 21:53.09 | pb_ | it's fairly meaningless to think of the upstream versions of hotplug, or glibc, or anything else as some kind of "gold standard". They just represent a snapshot of development at one particular point in time. |
| 21:53.37 | sgm|couchpotato | pb_: is this a known problem: http://pastebin.ca/5828 ? |
| 21:54.02 | sgm|couchpotato | configure.in exists. |
| 21:54.03 | treke | if it really bothers you that much, remove them. Otherwise, it's simpler for us to package all the scripts in one package |
| 21:54.28 | pb_ | sgm|couchpotato: I don't have a browser on hand, so I have no idea. |
| 21:54.37 | pb_ | if you can describe the problem in words, I can tell you. |
| 21:55.18 | reenoo_ | sgm|couchpotato: bk pull |
| 21:56.04 | sgm|couchpotato | pb_: will try reenoo_'s hint first. |
| 21:56.16 | pb_ | ok |
| 21:56.36 | reenoo_ | pb_: looks like the problems you fixed with your last two commits |
| 21:56.43 | pb_ | ah right |
| 21:56.51 | pb_ | in that case, yeah, bk pull is the correct answer :-) |
| 21:57.00 | sgm|couchpotato | pb_ did you change something in atk? |
| 21:57.12 | pb_ | no |
| 21:57.50 | sgm|couchpotato | pb_ my problem is that with atk 1.9.0 the do_configure failes with the message missing configure.ac |
| 21:57.58 | sgm|couchpotato | but there is an configure.in |
| 21:58.13 | sgm|couchpotato | trying last pull now |
| 22:01.06 | dave-osk | opie-collections says "if rootfs > 24 MB then add konq" ... |
| 22:01.23 | dave-osk | ... presumably because it assumes opie + conq fits in 24 MB ... |
| 22:01.48 | dave-osk | ... but the size in my build was 30 MB. So, shouldn't opie-collections check "30" not "24"? |
| 22:02.29 | sgm|couchpotato | reenoo_: pb_: OK, seems to work now. Tnx! |
| 22:03.22 | pb_ | very good |
| 22:03.40 | sgm|couchpotato | Is this a good practice in .bb files: |
| 22:03.46 | sgm|couchpotato | DEPENDS = "xproto x11 xextensions-${PV}" |
| 22:04.12 | sgm|couchpotato | that ${PV} stands for PV = "0.0cvs${CVSDATE}" |
| 22:04.28 | sgm|couchpotato | (this example is xext_cvs.bb) |
| 22:04.48 | pb_ | no, that's not terribly good practice. |
| 22:04.56 | pb_ | xextensions might have a different CVSDATE to the depending package. |
| 22:05.34 | pb_ | I don't think there's any better way to express it, though, so I guess we will have to live with that kind of thing for now. |
| 22:05.48 | sgm|couchpotato | pb_: what is better? I need a newer xentensions for xext since a few days. Actually, I need some hints in how to do this properly in the whole OE tree. |
| 22:06.21 | sgm|couchpotato | because I think mixing 1.0.1 and 0.0cvs(something) versions of xextensions does not seem a good idea |
| 22:09.58 | sgm|couchpotato | OMG... now gtk+ fails... |
| 22:11.11 | sgm|couchpotato | Cannot load module /home/sgm/work/OE/build/tmp/work/gtk+-2.6.2-r0/gtk+-2.6.2/modules/input/im-xim.la: /home/sgm/work/OE/build/tmp/work/gtk+-2.6.2-r0/gtk+-2.6.2/modules/input/.libs/im-xim.so: undefined symbol: XmbResetIC |
| 22:17.14 | RP__ | dave-osk: That'll be because konq is only installed on devices with a flash larger enough for it to fit onto. I think oe currently knows two flash sizes - less than 24MB and greater. If you rootfs can fit it on, it installs it... |
| 22:17.59 | treke | is it even that complicated? |
| 22:18.23 | treke | I'd imagine it's probably closer to an override for devices we want it on |
| 22:18.24 | RP__ | treke: maybe not. I haven't looked at the code... |
| 22:18.39 | mickeyl | look at opie-collections.inc |
| 22:18.47 | mickeyl | it checks for ROOT_FLASH_SIZE |
| 22:18.52 | mickeyl | with a bit of python code |
| 22:19.55 | beewoolie | I'm trying to build for openzaurus-3.5.3 and am getting an error : gzip: /a/trans/embedded/oe/sources/patch-2.4.18-rmk7.gz: No such file or directory |
| 22:20.24 | beewoolie | Can someone hook me up with a working local.conf to build for openzaurus? |
| 22:20.49 | mickeyl | this is pretty meaningless since local.conf is tied to a certain machine |
| 22:21.05 | *** join/#oe Tygger_Bob (~myname@64.26.144.98) |
| 22:21.05 | mickeyl | why not start with local.conf.sample and adjust it to your needs ? |
| 22:21.30 | treke | and pleaase don't build oe just for the sake of building it. It just makes everyone's lives more difficult |
| 22:24.10 | beewoolie | treke: for what I'm doing, the local.conf should be really simple. |
| 22:24.22 | beewoolie | The trouble is that it is really hard to get the combo right. |
| 22:24.39 | beewoolie | I'm not doing this for grins. I am trying to get a version of opie to build that works. |
| 22:24.44 | CIA-8 | 03mickeyl 07 * r1.3236.1.1 10openembedded/packages/ (4 files in 3 dirs): first shot at cleaning up the font situation in qte: split common qte fonts out of the libqte2 build and add qpf-qte-fonts |
| 22:25.17 | treke | if that's all you need, use one of the released images |
| 22:25.27 | treke | building your own is just a hassle. |
| 22:25.31 | beewoolie | What is difficult is getting all of the little pieces right. Not all distro/machine combinations work. AFAICT, there is no guideline for getting the combo right. |
| 22:25.41 | mickeyl | http://pastebin.com/241855 |
| 22:25.42 | beewoolie | I needto build because I need to hack it. |
| 22:25.43 | mickeyl | that's my site.conf |
| 22:26.02 | mickeyl | my local.conf depends on the machine I want to target and usually only has |
| 22:26.05 | mickeyl | MACHINE and DISTRO set |
| 22:27.17 | koen | I use auto.conf for MACHINE |
| 22:27.32 | koen | makes switching a lot easier |
| 22:27.33 | beewoolie | mickeyl: OK. What I've got as a baseline is now successfully building openslug. So, I believe it should work for others. Why would I get the error about the patch file? |
| 22:27.46 | mickeyl | what error ? |
| 22:27.52 | beewoolie | Scroll back. |
| 22:28.09 | beewoolie | 14:18 |
| 22:28.21 | koen | wow, 7 hours back ;) |
| 22:28.39 | treke | you didn't say anything at 14:18... |
| 22:28.53 | beewoolie | I'll paste it again. |
| 22:29.00 | beewoolie | <PROTECTED> |
| 22:29.04 | mickeyl | ah |
| 22:29.05 | mickeyl | no idea |
| 22:29.12 | beewoolie | Right. That's my point. |
| 22:29.12 | mickeyl | sounds like your bitbake is broken |
| 22:29.20 | beewoolie | It's up-to-date with svn. |
| 22:29.38 | mickeyl | did you do the usual steps? |
| 22:29.41 | mickeyl | cleaning the package |
| 22:29.42 | beewoolie | At revision 148. |
| 22:29.43 | mickeyl | building new |
| 22:29.51 | beewoolie | I deleted the tmp and started from scratch. |
| 22:29.51 | mickeyl | checking if there is a md5 but not a source |
| 22:29.52 | mickeyl | etc. |
| 22:30.16 | beewoolie | mickeyl: md5 of what? |
| 22:30.18 | mickeyl | is /a/trans/embedded/oe/sources/patch-2.4.18-rmk7.gz.md5 present ? |
| 22:30.26 | beewoolie | I've updated oe from BK. |
| 22:30.56 | mickeyl | ? |
| 22:31.04 | mickeyl | this is your sources dir |
| 22:31.09 | mickeyl | stuff gets downloaded there by oe |
| 22:31.15 | mickeyl | this is not updated with bk |
| 22:31.18 | beewoolie | Am I supposed to delete the sources directory too? |
| 22:31.26 | mickeyl | just answer my question |
| 22:31.36 | beewoolie | ...checking |
| 22:32.02 | mickeyl | when you kill oe in the middle of a fetch operation |
| 22:32.03 | beewoolie | Looks like, yeah, there is an md5 and no output file. |
| 22:32.09 | mickeyl | ha |
| 22:32.10 | mickeyl | delete it |
| 22:32.17 | mickeyl | then oe will see that the file is missing |
| 22:32.20 | mickeyl | and will redownload |
| 22:32.22 | beewoolie | So, are you saying that bb breaks and cannot fix this sort of problem? |
| 22:32.34 | mickeyl | that happens, when you interrupt oe |
| 22:32.37 | beewoolie | Has anyone written a script to check for this? |
| 22:32.41 | mickeyl | no |
| 22:32.45 | mickeyl | it rarely happens |
| 22:33.13 | beewoolie | You can say the same for any failed race condition. |
| 22:33.47 | mickeyl | well it doesn't seem to be a problem for most of us |
| 22:33.58 | mickeyl | or else someone would've changed oe |
| 22:34.09 | mickeyl | just don't interrupt oe when it fetches |
| 22:34.17 | beewoolie | nice try |
| 22:34.35 | RP__ | beewoolie: patches welcome... |
| 22:34.50 | beewoolie | RP__: I've submitted several... |
| 22:35.02 | beewoolie | RP__: that is not to say that they have been rejected. |
| 22:35.07 | beewoolie | ;-) |
| 22:35.20 | RP__ | beewoolie: excellent :) |
| 22:35.28 | mickeyl | to oe@handhelds.org ? |
| 22:35.43 | beewoolie | All it has to do is check that there is a download target when it fetches in addition to the md5 |
| 22:36.03 | beewoolie | I'm not much for python, so I stay out of that area. |
| 22:36.30 | koen | it would be nice to make the md5 a file actually containing the md5 |
| 22:36.35 | mickeyl | heh indeed |
| 22:36.42 | mickeyl | pester the guy who wrote the fetcher code |
| 22:36.44 | mickeyl | heh |
| 22:37.24 | mickeyl | i rather not have it computing and chekcing the md5, to be honest |
| 22:37.27 | mickeyl | it's slow enough without doing that |
| 22:37.30 | mickeyl | :D |
| 22:37.43 | CosmicPenguin | And if its corrupted, you'll find out soon enough |
| 22:37.51 | reenoo_ | heh |
| 22:37.57 | mickeyl | so in effect, we should kill of the md5 file |
| 22:38.00 | mickeyl | and look for the source file instead |
| 22:38.07 | mickeyl | kergoth`bbl, can you hear us ? |
| 22:38.09 | mickeyl | :D |
| 22:38.30 | reenoo_ | some sort of checksum should really be computed |
| 22:38.50 | reenoo_ | to check for backdoored tarballs and stuff |
| 22:39.33 | koen | so we'd have SRC_MD5=234852758375834758943 in a .bbb |
| 22:39.37 | koen | .bb* |
| 22:40.22 | reenoo_ | I'd also like .ipks to be gpg signed for the same reason but that's a different story |
| 22:40.23 | *** join/#oe keturn (~kevint@c-24-21-141-200.client.comcast.net) |
| 22:40.55 | reenoo_ | s/I'd/personally, I'd/ |
| 22:41.07 | koen | if someone would modify ipkg-upload..... |
| 22:42.25 | Pigi | nite all |
| 22:42.33 | koen | 'night Pigi |
| 22:43.15 | giel | so, how would one go about if i'd like to apply a patch only if i'm building against uclibc? |
| 22:43.44 | treke | we need the md5 file |
| 22:43.46 | treke | desperatly |
| 22:43.54 | treke | at least I do |
| 22:44.18 | sgm|sleepy | Q: what is the lag between the BK tree and the CVS one? |
| 22:44.21 | treke | I need some way of telling a file was successfuly download |
| 22:44.25 | treke | downloaded |
| 22:44.36 | treke | few hours? |
| 22:45.11 | *** join/#oe erich (~erich@h000625eb4eaf.ne.client2.attbi.com) |
| 22:45.16 | chouimat | <PROTECTED> |
| 22:46.06 | sgm|sleepy | treke: a few hours? They differ 123 files ATM. |
| 22:46.17 | treke | sgm|sleepy: so? |
| 22:46.26 | treke | a few hours is probably the case |
| 22:46.35 | treke | one hour at the absolute best case |
| 22:46.35 | reenoo_ | chouimat: heh |
| 22:46.43 | sgm|sleepy | treke: well, that's quite a lot of change in a few hours. |
| 22:46.49 | sgm|sleepy | s/change/changes |
| 22:46.58 | treke | that looks reasonable to me |
| 22:47.00 | *** join/#oe mithro (~tim@202.191.111.52) |
| 22:47.05 | sgm|sleepy | so I'm quite impressed |
| 22:47.22 | sgm|sleepy | and a bit worried too if I cannot use bk anymore |
| 22:47.41 | koen | sgm|sleepy: see http://openembedded.bkbits.net:8080/openembedded/index.html |
| 22:47.45 | mickeyl | just continue to use it ;) |
| 22:49.12 | sgm|sleepy | mickeyl: well, I now have an eval license because I cannot use the "free use" one. |
| 22:49.30 | treke | there are always the tarballs or cvs |
| 22:49.55 | *** join/#oe erich141 (~erich@h000625eb4eaf.ne.client2.attbi.com) |
| 22:50.00 | mickeyl | hey erich141 |
| 22:50.01 | sgm|sleepy | treke: I know. BTW. is there an automated way to update a snapshot with the patches? |
| 22:50.02 | mickeyl | long time no see |
| 22:50.25 | treke | sgm|sleepy: write a shell script to do it |
| 22:50.35 | erich141 | hi mickeyl, long time indeed ;-) |
| 22:50.41 | koen | sgm|sleepy: that's how the cvs snapshot works |
| 22:50.41 | treke | download tarball, store date of tarball |
| 22:50.48 | reenoo_ | hey erich141 |
| 22:50.53 | treke | in the future 1) download needed patches, store new date |
| 22:51.39 | erich141 | hi reenoo_ |
| 22:51.41 | treke | probably not too tough if you had the motivation to write it |
| 22:52.06 | treke | I've never had any needs to use the snapshots, so I never bothered trying to automate it |
| 22:52.56 | koen | treke: ask pb_, he uses the snaps for the cvs gateway iirc |
| 22:53.20 | sgm|sleepy | treke: CVS or SVN are more preferred IMO. Then you have a way to find your diffs in the oe dir. |
| 22:53.33 | treke | sgm|sleepy: then use the cvs |
| 22:54.50 | sgm|sleepy | treke: I will. but new users currently only have the choice between snapshots and BK because the cvs tree is not on the Wiki AFAIK. |
| 22:55.06 | treke | who care's if it's on the wiki? |
| 22:55.16 | sgm|sleepy | new users? |
| 22:56.26 | treke | easy solution is to update the wiki, I guress |
| 22:56.28 | treke | guess |
| 22:57.07 | sgm|sleepy | I already prosed it. But someone else (forgot who) would do it. But probably forgotten too. |
| 22:57.46 | hrw | mickeyl: qpf fonts from qt/e 2.3.10 needs more splitting based on their size |
| 22:58.48 | mickeyl | i know |
| 22:59.20 | treke | mickeyl: What are you thinking for 3.5.3? 2.6 or 2.4? |
| 22:59.29 | mickeyl | treke: I'm afraid 2.4 |
| 22:59.32 | hrw | in next days I will look on free ttf fonts in debian and generate qpf ones |
| 22:59.51 | mickeyl | treke: don't you think? |
| 23:00.02 | treke | we could probably get on fine just using the ttf's |
| 23:00.23 | mickeyl | treke|ho1e: I'd like to release both |
| 23:00.27 | treke | mickeyl: I dunno. I'd be inclined to do 2.6 on any plaftorm it works on |
| 23:00.29 | mickeyl | treke|home: so that the 2.6 image gets more testing |
| 23:00.39 | treke | mickeyl: and if it looks to be a problem, then release a 2.4 image |
| 23:00.51 | mickeyl | well |
| 23:00.53 | mickeyl | problems are: |
| 23:00.55 | mickeyl | irda |
| 23:00.57 | mickeyl | orinoco |
| 23:00.59 | mickeyl | sound |
| 23:00.59 | hrw | treke: probably - but where to store them? /usr/share/fonts/ for GPE or /opt/QtPalmtop/lib/fonts/ for OPIE? |
| 23:01.03 | mickeyl | leds |
| 23:01.23 | treke | hrw: /usr/share/fonts seems like the most appropriate place |
| 23:01.25 | mithro | orinoco? |
| 23:01.37 | mickeyl | 0.11e no longer builds against modern kernels |
| 23:01.43 | mickeyl | 0.15 is broken and no one can fix it |
| 23:01.45 | treke | why do we need 0.11e? |
| 23:01.48 | mickeyl | or wants to |
| 23:01.48 | hrw | mickeyl: what about oz 3.5.3/unstable with 2.4 and as a bonus some experimental 2.6 support? |
| 23:01.51 | erich141 | did anyone try TT/E with ttf support compiled in? |
| 23:01.54 | koen | mickeyl: release two sets of images: 2.4 and 2.6 ones |
| 23:02.01 | erich141 | s/TT/QT |
| 23:02.03 | hrw | erich141: I use ttf via opie-freetype |
| 23:02.25 | treke | hrw: the question is whether or not that is even necessary |
| 23:02.34 | treke | since binary compatibility isn't as much of a concern |
| 23:02.46 | mickeyl | indeed |
| 23:02.52 | mickeyl | we even could alter the qpe.conf now |
| 23:03.01 | mithro | mickeyl: why do we need the orinoco drivers? |
| 23:03.04 | mickeyl | since for all older stuff we have to use the compat libs anyway |
| 23:03.18 | mickeyl | mithro: because hostap doesn't work with hermes nor spectrum cards |
| 23:03.19 | koen | mickeyl: wlan? |
| 23:03.57 | mithro | mickeyl: ahh, i don't have one of thoses cards ;) |
| 23:06.28 | koen | 'night all |
| 23:06.29 | treke | mickeyl: and there is no new driver for them? |
| 23:07.32 | treke | http://www.nongnu.org/orinoco/ |
| 23:07.44 | treke | "The driver support Linux kernels from 2.4.10 to the latest 2.6 kernels (2.6.5 at the time of this release)." |
| 23:08.20 | mickeyl | the driver is broken |
| 23:08.22 | mickeyl | for arm |
| 23:08.29 | mickeyl | and no one is either capable |
| 23:08.29 | mickeyl | or interested |
| 23:08.30 | mickeyl | in fixing it |
| 23:08.32 | mickeyl | :/ |
| 23:08.57 | hrw | mickeyl: do oz 3.5.3 have planed release time? |
| 23:09.03 | mickeyl | somewhere between 0.11e and 0.15rc1 they broke it for arm |
| 23:09.19 | mickeyl | hrw: still mid-march if you ask me |
| 23:09.24 | hrw | mickeyl: ok |
| 23:10.03 | hrw | mickeyl: this time we really should release also sdk/x86 for those who cant build it using oe |
| 23:10.45 | hrw | mickeyl: I can take 'upgrades' feed after release - like I did recently with 3.5.2 |
| 23:11.26 | mickeyl | hrw: ok, thanks. that was good work |
| 23:11.32 | mickeyl | i'm pretty exhausted after every release |
| 23:11.49 | *** join/#oe _keith (~keith@bh02i525f01.au.ibm.com) |
| 23:12.09 | mickeyl | this time we also release opie 1.2.0 |
| 23:12.12 | mickeyl | which makes it worse |
| 23:12.24 | hrw | mickeyl: http://www.oesf.org/forums/index.php?showtopic=10775 - OZ users will try to unite and prepare some docs |
| 23:12.24 | mickeyl | aah |
| 23:12.54 | mickeyl | hrw: yes, i've read it. i won't be happy until something gets out of it though. there have been too much initiatives started and fade into nirvana lately |
| 23:13.04 | hrw | true |
| 23:15.44 | RP__ | mickeyl: I'd be tempted to go for a 2.6 kernel offically for c7x0 for the next oz release. The rootfs image is should now be compatible anyway... |
| 23:15.59 | RP__ | What's wrong with sound? |
| 23:16.13 | mickeyl | still that zombie on record |
| 23:16.15 | CIA-8 | 03mickeyl 07 * r1.3243 10openembedded/packages/qpf-fonts/qpf-qte_2.3.10.bb: make qpf-qte no longer suck thanks to the usage of do_split_packages |
| 23:16.19 | mickeyl | no, i didn't have time to test against 2.4 yet |
| 23:16.27 | RP__ | and I can put something together for led support... |
| 23:16.34 | RP__ | mickeyl: ah, yes :-/ |
| 23:17.04 | mickeyl | lets watch how things turn out in the next 4 weeks |
| 23:17.09 | mickeyl | we can decide 1 week before |
| 23:17.14 | mickeyl | and then start to polish |
| 23:17.23 | RP__ | mickeyl: Ok, fine by me |
| 23:17.25 | mickeyl | after all, this is still an unstable tree |
| 23:17.34 | hrw | I can write in Polish ;) |
| 23:17.39 | mickeyl | heh |
| 23:17.41 | RP__ | I'll continue to try and improve corgi 2.6... |
| 23:17.56 | mickeyl | you've doing a helluva job. did i say that lately yet ? |
| 23:17.57 | mickeyl | :) |
| 23:18.12 | RP__ | I'm getting caught up in a lot of work on the ati driver atm :-/ |
| 23:18.23 | RP__ | thanks :) |
| 23:18.34 | *** join/#oe jamie (~jamie@modem-980.porcupine.dialup.pol.co.uk) |
| 23:18.59 | mickeyl | heh, but when you get that w100 to cooperate, we have eliminated one major reason not to go OZ |
| 23:19.09 | mickeyl | nearly the 2nd last obstacle |
| 23:19.38 | mickeyl | the last - being not compatible - which we won't tackle anyway |
| 23:19.55 | RP__ | Well, it works as it stands. There are just more things we can make it do... |
| 23:20.16 | RP__ | I was pleased to see someone trying to accelerate qpe |
| 23:20.32 | mickeyl | yep, sirfreds work may be pretty useful for qte |
| 23:20.38 | erich141 | does new 2.6 kernel has basic functionality support for poodle? |
| 23:21.09 | RP__ | erich141: depends what you mean by basic. The biggest problem is no frontlight |
| 23:21.25 | hrw | bye |
| 23:21.25 | RP__ | The image in oe will compile and give you a serial console |
| 23:21.31 | mickeyl | c'ya hrw |
| 23:21.32 | hrw | good night/day |
| 23:21.36 | RP__ | ttfn hrw |
| 23:21.42 | hrw | ttfn? |
| 23:21.46 | erich141 | RP__: as long as it boots it has basic support in my opinion :) |
| 23:21.52 | RP__ | hrw: bye :) |
| 23:21.56 | mickeyl | yep, core machine support is there |
| 23:22.06 | mickeyl | now missing are "just" drivers |
| 23:22.15 | RP__ | (ta-ta for now) |
| 23:22.23 | erich141 | night hrw |
| 23:22.29 | RP__ | erich141: it boots |
| 23:22.50 | erich141 | and there is no chance to get SD working in 2.6? |
| 23:22.55 | RP__ | work on locomo support of the back/frontlight whatever it has would be nice |
| 23:22.59 | mickeyl | SD is no problem on everything but collie |
| 23:23.07 | RP__ | erich141: SD works on the corgis |
| 23:23.10 | erich141 | that's good news |
| 23:23.13 | mickeyl | indeed |
| 23:23.14 | RP__ | and should be easy enough on poodle |
| 23:23.25 | mickeyl | just not for those poor 5500 / 5000 weenies |
| 23:23.30 | erich141 | you guys did tremendous job in the last few months |
| 23:23.30 | mickeyl | (including me) |
| 23:24.18 | mickeyl | thanks. the long stainy tedious and boring work on the build system finally pays off |
| 23:24.20 | Skyhusker|bed | hey mickeyl |
| 23:24.26 | mickeyl | hi Skyhusker|bed |
| 23:24.46 | Skyhusker|bed | i haven't get a bt yet, but i isolated the line which caused the segfault (in opienetwork) |
| 23:24.52 | RP__ | mickeyl: The build system is looking good here. I'm very impressed with it :) |
| 23:25.09 | mickeyl | Skyhusker|bed: ok, that's good enough for me to check |
| 23:25.13 | mickeyl | RP: :) |
| 23:25.42 | Skyhusker|bed | at line 931 in my source file, maybe a bit different in the one in cvs |
| 23:25.44 | Skyhusker|bed | stations->last()->ssid = static_cast<const char*>( we->u.essid.pointer ); |
| 23:26.01 | Skyhusker|bed | i think it is because we->u.essid.pointer is not NULL terminated |
| 23:26.28 | mickeyl | didn't I insert a \0 at we->u.essid.len before ? |
| 23:26.45 | Skyhusker|bed | case SIOCGIWESSID: |
| 23:26.45 | Skyhusker|bed | <PROTECTED> |
| 23:26.46 | Skyhusker|bed | <PROTECTED> |
| 23:26.46 | Skyhusker|bed | <PROTECTED> |
| 23:26.46 | Skyhusker|bed | <PROTECTED> |
| 23:26.46 | Skyhusker|bed | <PROTECTED> |
| 23:27.00 | RP__ | mickeyl: I like the new c7x0. It makes a lot of sense. |
| 23:27.00 | mickeyl | hmm |
| 23:27.08 | Skyhusker|bed | i tryed memcpy'ing the string and adding a \0, but didn't worked anyway |
| 23:27.18 | mickeyl | RP__: yeah, it was overdue. i don't get why we didn't do it before :) |
| 23:27.28 | mickeyl | Skyhusker|bed: indeed. that's right. i'll fix that. thanks |
| 23:27.35 | Skyhusker|bed | mickeyl: thanks to you ;) |
| 23:27.41 | Skyhusker|bed | time to go bed now |
| 23:27.45 | mickeyl | have a goodnite |
| 23:27.46 | Skyhusker|bed | good night all |
| 23:27.49 | Skyhusker|bed | thanks :D |
| 23:28.46 | treke | mickeyl: I think it was so that we didn't have to build three kernels each time we wanted to test something :) |
| 23:30.33 | mickeyl | treke: hehe, yeah maybe that was it. |
| 23:30.46 | mickeyl | although it's still possible to build just one kernel now |
| 23:30.49 | treke | at this point the 2.4 kernel will probably never be touched again |
| 23:30.55 | mickeyl | let's hope |
| 23:30.55 | mickeyl | heh |
| 23:31.12 | treke | except maybe for spitz |
| 23:31.15 | RP__ | I keep a copy of its source around. Does that count? :) |
| 23:31.22 | mickeyl | RP__: hardly ;) |
| 23:31.27 | beewoolie | mickeyl: I've got a simple question. Your local.conf says that you provide gcc2.95 for arm, right? |
| 23:31.34 | treke | RP likes to have some scary reading before he goes to bed |
| 23:31.35 | mickeyl | beewoolie: yes, i do |
| 23:31.52 | mickeyl | beewoolie: i have a tweaked copy of an old 2.95.3 laying around, |
| 23:31.56 | mickeyl | in my PATH |
| 23:31.58 | beewoolie | In the build of /openzaurus-pxa-2 |
| 23:31.58 | beewoolie | .4.18-rmk7-pxa3-embedix-r32 |
| 23:32.05 | beewoolie | I'm seeing a failure. |
| 23:32.19 | treke | probably because you don't have gcc 2.95 in your path |
| 23:32.22 | mickeyl | have a look at ZaurusKernels on the wiki. It will tell you that you have to rename two files |
| 23:32.28 | beewoolie | I'm not sure if it is failing to find the compiler, or the file it is compiling. |
| 23:32.37 | beewoolie | arm-linux-gcc-2.95: No such file or directory |
| 23:32.44 | mickeyl | yep. read ZaurusKernels |
| 23:32.53 | beewoolie | The reason I ask is that I don't provide the compiler. |
| 23:32.57 | mickeyl | uhm |
| 23:32.59 | treke | We know |
| 23:32.59 | mickeyl | you need to |
| 23:33.02 | treke | READ THE WIKI |
| 23:33.08 | beewoolie | Ah. |
| 23:33.12 | mickeyl | there's no way, oe will build you a 2.95.3 |
| 23:33.16 | mickeyl | (unfortunately) |
| 23:33.22 | beewoolie | OK. I've already got it on here. |
| 23:33.30 | beewoolie | thx |
| 23:36.18 | *** join/#oe pb_ (~pb@p43.n-nypop05.stsn.com) |
| 23:36.54 | RP__ | treke: The 2.4 kernel sends shivers down my spine... |
| 23:37.06 | beewoolie | RP__: Me too. me too. |
| 23:39.56 | reenoo_ | heh |
| 23:42.05 | CIA-8 | 03mickeyl 07 * r1.3245 10openembedded/ (6 files in 5 dirs): catch up with new default fonts and remove the beagle-specific hack which is no longer needed because ~root is /home/root on all models now |
| 23:42.10 | CIA-8 | 03mickeyl 07 * r1.3244 10openembedded/packages/meta/opie-collections.inc: opie-collections: consolidate c7x0 and catch up w/ more fine granular font packaging |
| 23:42.25 | mickeyl | ok, lets do a opie-image from scratch to see what i broke |
| 23:42.27 | mickeyl | g'night folks! |
| 23:42.35 | reenoo_ | 'night mickey|zzZZzz |
| 23:42.54 | RP__ | 'night mickey|zzZZzz |
