**** BEGIN LOGGING AT Fri Jun 24 23:59:57 2005 Jun 25 00:24:02 i am unable to download oe Jun 25 00:30:22 when i run bitbake bootstrap-image and am getting the following error please see http://pastebin.ca/15769 Jun 25 00:31:56 sounds like a problem with your environment or with savannah. any firewalls and blocked ports on your site? can you reach savannah via ping? Jun 25 00:32:41 hmm savannah seems to be offline Jun 25 00:33:45 we don't control upstream sites Jun 25 00:33:55 try grabbing the files from a source mirror Jun 25 00:33:59 unable to ping savannah ? Jun 25 00:33:59 i.e. vanille.de/sources/ Jun 25 00:35:24 indeed, take a look at the kde buildsystem Jun 25 00:35:27 aargh Jun 25 00:35:30 heh Jun 25 00:35:32 good morning Jun 25 00:35:34 confusing your # again? :) Jun 25 00:35:52 no, irssii + pastebuffers Jun 25 00:37:53 If it keeps raining and thundering this way I might even add gnome 2.10 to OE Jun 25 00:38:14 heh Jun 25 00:38:15 sounds good Jun 25 00:38:53 I allready patched libtool for it.... Jun 25 00:38:58 ~lart libtool regressions Jun 25 00:38:58 * ibot holds libtool regressions to the floor and spanks him with a cat-o-nine-tails Jun 25 00:39:47 i had the 770 in my hands on thursday, i can't wait to own one. it feels solid and professional and is shirt pocket compatible Jun 25 00:40:03 a bit sluggish, admittedly Jun 25 00:40:08 but good enough Jun 25 00:40:24 they still use hardfloat Jun 25 00:40:29 aah Jun 25 00:40:30 so it should be eas to improve it Jun 25 00:40:48 and we still have RP :) Jun 25 00:41:03 what's your position regarding our merger now? Are you still for doing it and letting the veterans take over familiar? Jun 25 00:41:15 yeah, ~praise RP just because Jun 25 00:41:51 Jamey's fedora thing sounded good to me Jun 25 00:42:41 we merge and and stay bleeding edgge, and familiar becomes like debian stable Jun 25 00:43:24 yeah, that sounds like a good way. Jun 25 00:43:32 that also means we can vote on one name less :) Jun 25 00:43:37 :) Jun 25 00:44:37 that only leaves 3 (not counting SMILE) Jun 25 00:44:55 right. the con of SMILE is convincing Jun 25 00:50:39 ok, page adjusted Jun 25 00:51:01 * koen still votes for Ã¥ngström Jun 25 00:51:33 I agree Jun 25 00:51:55 Ângstrom would be just cool Jun 25 00:52:01 if he were finnish we could have asked for nokia funding ;) Jun 25 00:52:17 given that our two votes for sure count the most, I don't think there's any point in voting at all ;) Jun 25 00:52:26 ;) Jun 25 01:03:36 MOrning Jun 25 01:03:51 hey uv1 Jun 25 01:04:15 wow. it takes a while to pull the monotone repository for monotone. can only imagine how long it will take on the OE or kernel repository in a year or two from now. Jun 25 01:05:38 monotone will hopefully have advanced until then Jun 25 01:06:39 PARALLEL_MAKE strike again Jun 25 01:06:46 ~lart espgs-native Jun 25 01:06:46 * ibot grabs a large, mis-shapened log, with squirrels, and beats espgs-native until only the nuts remain ... which the squirrels run off with Jun 25 01:09:38 * koen joins #monotone Jun 25 01:16:12 03koen 07 * r1.3615 10openembedded/packages/espgs/espgs_7.07.1rc1.bb: espgs_7.07.1rc1.bb: disable parallel make Jun 25 01:16:14 03koen 07 * r1.3612.1.2 10openembedded/packages/espgs/espgs-native_7.07.1rc1.bb: espgs-native_7.07.1rc1.bb: disable parallel make Jun 25 01:16:17 03koen 07 * r1.3612.1.1 10openembedded/packages/gpe-mini-browser/gpe-mini-browser_0.11.bb: Add gpe-mini-browser, courtesy Phillipe DeSwert Jun 25 01:17:25 I like it when this happens: Jun 25 01:17:26 10:16 < njs`> koen: annodomini here just finished a patch for it this evening, though, so should be any moment now :-) Jun 25 01:19:58 excellent Jun 25 01:20:20 huhu mickeyl Jun 25 01:21:09 hi woglinde Jun 25 01:21:19 * mickeyl disppointed Jun 25 01:21:24 i wanted to go playing squash Jun 25 01:21:33 cvs and ssh works on my system and bitbake bootstrap-image does not work ?. please see http://pastebin.ca/15769 Jun 25 01:21:37 what can i do ? Jun 25 01:21:44 i can't, since I walked too much yesterday and now my feet are aching :( Jun 25 01:24:58 i cant find gnu-config-native-0.1cvs20050625-r3 in vanilla.de Jun 25 01:26:20 well, this mirror runs not daily. grab a slightly older one and rename it or change your CVSDATE Jun 25 01:27:53 Could kaffe run on arm? like H5400? Jun 25 01:28:53 mickeyl: we're seeing some nslu2-linux problems with monotone. We may have to use an org.openembedded.nslu2-linux branch instead of the org.nslu2-linux.openembedded branch. Jun 25 01:29:56 it seems that if we commit a change in the org.nslu2-linux.openembedded branch, and then propagate it to org.openembedded.dev, it doesn't show up in your repo. We believe this is because you are not serving (and therefore not accepting from pushes) any branches other than org.openembedded.* (which is the right thing to do). Jun 25 01:30:26 so it looks like our nslu2-linux branch will need to be under org.openembedded.* for us to be able to propagate changes back Jun 25 01:30:35 any one have any idea on "fastest jvm" availiable on embedded device? Jun 25 01:31:09 none Jun 25 01:31:16 :) Jun 25 01:31:24 gcc Jun 25 01:31:25 :) Jun 25 01:31:29 mickeyl: I'm going to test that theory in a couple of hours time by creating the org.openembedded.nslu2-linux branch. So I'm asking for our (nslu2-linux's) exclusive use of that branch name in your hierarchy Jun 25 01:31:49 rwhitby-away: ok, i see. yes, that'd work for us as well Jun 25 01:32:01 cool, thx. Jun 25 01:32:16 back in a couple of hours ... Jun 25 01:32:19 good luck Jun 25 01:48:26 hmm, tiff is missing in OE Jun 25 01:52:03 03koen 07 * r1.3616 10openembedded/packages/ (libtiff/tiff_3.7.2.bb poppler/poppler_0.3.3.bb): Jun 25 01:52:03 tiff_3.7.2.bb: new file Jun 25 01:52:03 poppler_0.3.3.bb: fix staging Jun 25 02:00:55 koen: can you pull from mtn.nslu2-linux.org/org.nslu2-linux ? Jun 25 02:01:12 check the viewmtn head of org.openembedded.dev before and after the pull Jun 25 02:01:34 (pull it into the ewi public repo that is) Jun 25 02:01:52 ~botmail for Xora: NOTE: package evince-0.3.2-r0: task do_build: completed Jun 25 02:05:45 koen: my theory is that we haven't been able to push stuff into org.openembedded.dev because the changes we are pushing were originally commited in org.nslu2-linux.openembedded, and so your server drops them when we push. If you pull, and those changes suddenly appear, then that would support the theory. Jun 25 02:06:18 it's pulling in 153 certs right now Jun 25 02:06:46 you did propagate and merge them into org.openembedded.dev? Jun 25 02:07:54 rwhitby-away: http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=051fc6f9e28e16d274e2cd95db273eb514d8fb33 Jun 25 02:08:08 koen: yes, we did Jun 25 02:08:37 ooh.. nice green color ;) Jun 25 02:09:09 ok, I think that confirms the theory. Jun 25 02:10:46 koen: that means that either 1) your server serves up just "org.*" instead of "org.openembedded.*" or 2) we rename our branch from org.nslu2-linux.openembededd to org.openembedded.nslu2-linux Jun 25 02:11:01 I think 2) is the more appropriate response (even though it means more work for me) Jun 25 02:11:53 and what it means for you is that any revision that is ever going to be propagated to an org.openembedded.* branch must also originate in an org.openembedded.* branch. Jun 25 02:11:57 ewww, nasty green color Jun 25 02:12:08 eeeyh Jun 25 02:12:15 It's MY green color! :P Jun 25 02:12:48 hmm, viewmtn looks interesting Jun 25 02:14:00 I can't seem to pull in those changes into dominion Jun 25 02:19:39 koen: what does heads on oe.dev say? Jun 25 02:19:51 (on ewi) Jun 25 02:20:45 monotone: branch 'org.openembedded.dev' is currently merged: Jun 25 02:20:45 e939e179c1552200dcb9eb26b181ac051f11e4be rwhitby@nslu2-linux.org 2005-06-25T06:35:15 Jun 25 02:21:29 did ewi fall down? Jun 25 02:22:20 koen: ok, that matches my repo Jun 25 02:22:22 no, I have to stop serving to check heads Jun 25 02:22:28 ah, heh Jun 25 02:22:38 koen: I don't think you have to stop it Jun 25 02:23:00 there is a note on data integrity Jun 25 02:23:16 i am facing an error while doing bitbake bootstrap-image Jun 25 02:23:40 koen: really? I missed that one. reference? Jun 25 02:23:56 http://pastebin.ca/15834 Jun 25 02:24:00 the manual iirc Jun 25 02:24:08 the note on "why not a real db" Jun 25 02:24:19 getting an error: please refer http://pastebin.ca/15834 Jun 25 02:24:36 while doing bitbake bootstrap-image Jun 25 02:26:31 did gnu-config download ok? Jun 25 02:27:42 1 sec. will be back... Jun 25 02:37:19 koen: did u go through my pastebin.ca Jun 25 02:37:26 i could not compile. Jun 25 02:37:30 yes Jun 25 02:37:39 did gnu-config download OK? Jun 25 02:38:00 where should i check ? Jun 25 02:38:21 ${DL_DIR} Jun 25 02:40:03 it pointing to home dir and nothing is there ? Jun 25 02:40:33 does ftp require passive support enabled in firewall ?. Jun 25 02:40:50 is there any way we can bypass this by setting an environment variable ? Jun 25 02:41:30 not sure if it has downloaded gnu-config ? Jun 25 02:41:32 there should be a DL_DIR setting in local.con Jun 25 02:42:17 gnu-config is not downloaded ? Jun 25 02:42:20 still Jun 25 02:42:37 what about ftp - passive support ? Jun 25 02:42:53 it is neccessary or is there anyway we can bypass it ? Jun 25 02:42:59 no idea Jun 25 02:43:00 during downloading of files ? Jun 25 02:43:34 koen: from the SQLite feature list: "Most SQL database engines are client/server based. Of those that are serverless, SQLite is the only one that this author knows of that allows multiple applications to access the same database at the same time. " Jun 25 02:43:40 where can i find the download script ? Jun 25 02:44:17 rwhitby-away: ok, that's good to know Jun 25 02:44:19 koen: and I read "monotone prohibits concurrent access to a single repo for reasons of simplicity, robustness, and confidence in correctness. We agree that it is occasionally annoying that access to the database is serialized" as saying that it is safe. Jun 25 02:44:57 koen: since the db is a single file, I expect it's just simple locking per transaction. Jun 25 02:45:08 (and monotone does everything in atomic transactions) Jun 25 02:46:24 ram123: could you set CVSDATE=20050619 in local.conf? Jun 25 02:51:16 done Jun 25 02:55:03 morning all Jun 25 02:56:57 morning Jun 25 02:57:41 but it is the ftp issue. Jun 25 02:58:16 the building process opens PASSV FTP connection and the server refuses that ? Jun 25 03:05:06 do u have all the sources at oesources.org for all the pakages. Jun 25 03:08:03 Good morning. Jun 25 03:08:18 morning sajar Jun 25 03:08:24 oops, morning SirFred Jun 25 03:08:43 rwhitby: :) Jun 25 03:09:19 ram123: you could try 'bitbake -c clean gnu-config-native' and deleting the it from your DL_DIR Jun 25 03:09:30 mickeyl: Perhaps you could help me with a problem. Jun 25 03:27:47 mickeyl: ping Jun 25 03:27:51 mickeyl: Don't hide. Jun 25 03:28:23 Well, I'm going to launch my question anyway, perhaps somebody could help. Jun 25 03:29:07 SirFred: I doubt mickeyl is "hiding". He's be idle for a couple of hours Jun 25 03:29:22 The problem is that once the c7x0 screen is rotated, new qpe clients like to provide a "Transformed:Rotx:0" screen spec instead of the original one (the one in the QWS_DISPLAY environment variable). Jun 25 03:29:29 NAiL: I was just joking. Jun 25 03:30:22 I've found that AppLnk::execute(const QStringList& args) seemed to be the place where that transformation was made. So I changed it to use the original driver but without luck. Jun 25 03:30:37 (AppLnk::execute is part of the libqpe code) Jun 25 03:32:20 Does anybody knows where that transformation is made? I mean, who and where takes the decission to change the QWS_DISPLAY environment when a new application is launched. Jun 25 03:35:10 Humm, I think I've found it. It was hidden inlined in a .h file. Jun 25 03:37:17 hi Jun 25 03:55:11 can we disable docs while building ? Jun 25 03:55:24 i dont want docs for any of them Jun 25 03:56:04 docs get installed in different packages, so they wont end up on the rootfs Jun 25 03:56:25 i am having trouble downloading the docs ? Jun 25 03:56:44 required packages and installing them so, i want to disable it. Jun 25 03:56:47 can i do that ? Jun 25 03:58:00 I don't really know what you're trying to do Jun 25 03:59:22 are you Jun 25 03:59:41 a) having trouble building packages Jun 25 03:59:55 b) don't want -doc packages in the rootfs Jun 25 04:00:07 c) don't want -doc packages at all Jun 25 04:15:50 koen|afk: what 'afk' means? Jun 25 04:15:52 hmmm Jun 25 04:15:59 away from keyboard Jun 25 04:16:15 | arm-linux-gcc: Internal error: Killed (program cc1) Jun 25 04:16:15 | Please submit a full bug report. Jun 25 04:16:30 xf86Bus.o killed gcc :) Jun 25 04:17:00 rwhitby: Thanks. Jun 25 04:18:24 Crofton|laptop: do you have enough memory? That looks like the OOM slayer Jun 25 04:18:35 mem/swap Jun 25 04:18:47 tip for young players: if you have previously deleted your monotone database, and lost your private key, then generated and distributed new keys, and created a new database, then don't delete that one too before you save the private key somewhere else .... Jun 25 04:19:00 * NAiL rolls over Jun 25 04:19:14 Mem: 1035112 1021384 13728 0 311744 369508 Jun 25 04:19:14 -/+ buffers/cache: 340132 694980 Jun 25 04:19:14 Swap: 1048568 511624 536944 Jun 25 04:19:48 Crofton|laptop: gcc *got* killed. Aka, something sent gcc a signal that asked it less than politely to bugger off Jun 25 04:20:06 but looks like you've got enough mem though ;) Jun 25 04:20:23 heh Jun 25 04:20:37 looks like it is gcc-cross-3.4.2-r10 Jun 25 04:20:52 er 4.4.3 Jun 25 04:20:59 3.4.3 bugger all Jun 25 04:21:53 * NAiL tries to fall asleep again. Jun 25 04:27:43 what do you guys think about adding "fuser" to the stock OZ ROMs? It would allow a clean umount of SD and CF cards during a reboot. Currently this doesn't happen. Jun 25 04:56:37 mickeyl: you guys are going to create a completely new monotone database on the changeover date, right? Jun 25 04:56:53 (as opposed to just propagating data from the current testing database ...) Jun 25 05:05:02 another tip for young players. replacing a pubkey in a monotone database is a very bad idea, and will often lead to having to completely redo the database from scratch. Jun 25 05:13:19 ~lart sdcontrol Jun 25 05:13:19 * ibot nabs the moon and broadsides sdcontrol with the sea of tranquility Jun 25 05:23:45 koen|afk: ping Jun 25 05:53:57 i am getting ERROR: Unable to parse conf/bitbake.conf Jun 25 05:54:02 what could be wrong ? Jun 25 05:54:18 wrong directory? Jun 25 05:54:33 what is the solution ? Jun 25 05:54:44 going to the rigth directory? Jun 25 05:54:51 koen ? Jun 25 05:58:30 BBPATH... Jun 25 05:58:41 he/she is gone Jun 25 05:58:54 03CoreDump 07 * r1.3617 10openembedded/packages/sharp-binary-only/sharp-sdmmc-support-2.4.18-rmk7-pxa3-embedix/ (sd sdcontrol): Jun 25 05:58:54 sdcontrol: Make 'kill_tasks' functional again and applied work-around for b0rked Jun 25 05:58:54 /proc/partitions from ZUG (credits go to ironstorm). /etc/init.d/sd: Use Jun 25 05:58:54 sdcontrol to umount all SD partitions. In combination with an installed 'fuser' Jun 25 05:58:54 this allows all partition to umount cleanly on a reboot. Jun 25 05:59:46 03CoreDump 07 * r1.3618 10openembedded/packages/sharp-binary-only/sharp-sdmmc-support_2.4.18-rmk7-pxa3-embedix.bb: Bump PR (sigh) Jun 25 06:20:26 re Jun 25 06:20:30 CoreDump|home: yes, I Jun 25 06:20:39 I agree. I want fuser and setserial in the base image Jun 25 06:20:57 rwhitby: yeah, i think we'll generate a new db on 1st of july. Jun 25 06:21:11 SirFred: pong Jun 25 06:21:16 mickeyl: good, that means I can regen my key with impunity Jun 25 06:21:21 sure :) Jun 25 06:22:00 CoreDump|home: is the busybox situation cleaned up or are we still suffering from that patch going wild? Jun 25 06:22:41 getting the error ERROR: Unable to parse conf/bitbake.conf Jun 25 06:22:43 eventhough my BBPATH points to correct place ? Jun 25 06:22:44 what to do ? Jun 25 06:23:02 possibly some errors in local.conf Jun 25 06:23:09 like ? Jun 25 06:23:14 hypens in machine name Jun 25 06:23:22 syntax errors Jun 25 06:23:25 ... Jun 25 06:31:08 mickeyl: the patch is still in, so I'd think the problem hasn't been fixed, yet Jun 25 06:31:16 i see Jun 25 06:31:25 what was it supposed to fix anyway? Jun 25 06:31:35 dunno :) ask pb Jun 25 06:31:51 file://celf-ash-builtins.patch;patch=1 \ Jun 25 06:31:58 that's the bugger Jun 25 06:32:06 ahah Jun 25 06:32:10 let me guess Jun 25 06:32:14 tests return value is going bogus? Jun 25 06:32:52 busysbox's "test" is fubared, always returning true Jun 25 06:32:55 yep Jun 25 06:33:10 where can I send a key for monotone access to? Jun 25 06:33:17 or can I do anonymous check out? Jun 25 06:33:18 zecke: atm., koen|afk is the authority Jun 25 06:33:25 where is koens blog again? Jun 25 06:33:59 zecke: you can anon pull from ewi546.ewi.utwente.nl org.openembedded Jun 25 06:34:19 then checkout the org.openembedded.dev branch Jun 25 06:34:41 CoreDump|home: from the look of it i'd suspect that to affect all platforms Jun 25 06:35:22 ah, i see. The bug is confirmed for collie and poodle FWIW Jun 25 06:35:41 rwhitby: version 18 is sufficent? Jun 25 06:35:57 OT: man evince is bad Jun 25 06:36:12 no Jun 25 06:36:16 monotone 18 is sluggish Jun 25 06:36:18 hi all Jun 25 06:36:20 you really want to use 0.19 Jun 25 06:36:23 hi bluelightning Jun 25 06:36:30 hey zecke, mickeyl Jun 25 06:37:04 mickeyl: weird I've have .19 Jun 25 06:37:44 mickeyl: Hi. Jun 25 06:37:53 mickeyl: Have you seen my patch? Jun 25 06:38:04 SirFred: just got your mail. your patch seems reasonable but lacks bootstrapping support from ODevice Jun 25 06:38:27 mickeyl: sure it lacks it, whatever it is. Jun 25 06:38:32 hehe Jun 25 06:38:35 mickeyl: :-P Jun 25 06:38:42 mickeyl: I don't know what is that supposed to be. Jun 25 06:38:45 ok, think about starting opie Jun 25 06:38:56 there is no QWS_DISPLAY set in that case Jun 25 06:39:04 so we need to put some more intelligence in the launcher Jun 25 06:39:07 or in odevice Jun 25 06:39:11 (which would be the better place) Jun 25 06:39:21 Well, isn't /etc/init.d/opie setting QWS_DISPLAY? Jun 25 06:39:28 no, thank god Jun 25 06:39:33 i did a lot of work Jun 25 06:39:37 for it not to be set :) Jun 25 06:39:49 mickeyl: So, I'm using an old version. Let me see... Jun 25 06:40:07 i have just given bitbake bootstrap-image how long it takes and what does it do ? Jun 25 06:40:14 mickeyl: Correct. Current opie doesn't set it. Jun 25 06:40:27 hold on I'll direct you to the appropriate place in the launcher Jun 25 06:40:49 http://handhelds.org/cgi-bin/cvsweb.cgi/opie/core/launcher/main.cpp Jun 25 06:40:57 see r1.43 Jun 25 06:40:59 mickeyl: does monotone scripting look like python? Jun 25 06:41:07 zecke: a bit. it's lua, in fat Jun 25 06:41:08 fact Jun 25 06:41:15 mickeyl: So are talking about ODevice and subclasses? Jun 25 06:41:42 oh Jun 25 06:41:42 SirFred: exactly. ODevice already knows about the default rotation hence it will be natural for it to know about the default output driver Jun 25 06:41:44 lua Jun 25 06:42:10 mickeyl: have you guys found a CIA client for monotone that runs on commit yet? The best we have found is one that runs from cron. Jun 25 06:42:11 what does bitbake bootstrap-image do ? Jun 25 06:42:51 rwhitby: writing a hook should not be too hard Jun 25 06:43:05 rwhitby: not yet, koen has briefly looked for one but didn't find one. Jun 25 06:43:33 mickeyl: I suppose the thing is about Transformation Zaurus::rotation() and it's usage in Zaurus::init Jun 25 06:44:16 zecke: note_commit would work as long as every developer is running the CIA client. There needs to be a note_netsync_commit or something to catch things from devlopers on push to the official repo. Jun 25 06:44:26 Perhaps the idea is to create a new member like const QString& odevice::qteDriver( void ) const or something so? Jun 25 06:44:47 SirFred: exactly like that Jun 25 06:45:01 the default implementation would return Transformed Jun 25 06:45:03 A virtual method, just using Transformed in the default one. Jun 25 06:45:09 *nod* Jun 25 06:45:10 mickeyl: :) Jun 25 06:45:42 mickeyl: and just use it in the specialized init() member. Jun 25 06:45:58 mickeyl: Perhaps I could be able to do it. Jun 25 06:46:20 Launcher::initEnvironment() Jun 25 06:46:22 * zecke reads virtual... and propagates usage of the virtual hook Jun 25 06:46:53 * mickeyl pretends not to hear Jun 25 06:47:18 * SirFred thinks 'bah, bitbaker jokes...' Jun 25 06:47:39 the thing is I broke compatibility in libopiecore over the last weeks with adding the new OPcmcia subsystem wrapper Jun 25 06:47:42 * CoreDump|home adds setserial & fuser to task-bootstrap, lacking a better place Jun 25 06:47:52 CoreDump|home: *nod* Jun 25 06:47:53 mickeyl: Where is that, is it part of the opie-taskbar? Jun 25 06:48:15 SirFred: yes. opie-taskbar ~= launcher Jun 25 06:48:22 mickeyl: What a lot of patches. Jun 25 06:48:23 * mickeyl watches an impressive downpour Jun 25 06:48:37 * SirFred looks into a dictionary for downpour Jun 25 06:48:38 pb_: ping Jun 25 06:48:52 mickeyl: at your service Jun 25 06:48:59 hello what does bitbake boot-strap do ? Jun 25 06:49:02 pb_: heya. what can we do about the celf patch breaking bb's test applet? Jun 25 06:49:07 * rwhitby creates the org.openembedded.nslu2-linux branch Jun 25 06:49:22 SirFred: just a more impressive word for a rapid rain shour Jun 25 06:49:29 * mickeyl works on his literaric english Jun 25 06:49:41 * SirFred curses babelfish. Jun 25 06:49:58 * SirFred needs to buy a good dictionary Jun 25 06:50:02 ~praise dict.leo.org Jun 25 06:50:02 All hail dict.leo.org! Jun 25 06:50:04 mickeyl: Thanks. Jun 25 06:50:33 mickeyl: back out the patch, I suppose. I asked the bug submitter whether turning off CONFIG_FEATURE_BUILTIN_TEST fixed it, but he didn't reply. Jun 25 06:50:35 English-German, how useful for me! Jun 25 06:50:39 ;-) Jun 25 06:51:15 I will be able to search how unknow english words are written in an completely unknown language Jun 25 06:51:17 pb_: ok. since there's no talk about the patch in the last two months of busybox traffic it looks like they're not even considering it Jun 25 06:51:24 SirFred: oops ;) Jun 25 06:51:42 * mickeyl disables it for now Jun 25 06:51:47 mickeyl: So, I need to write two new patches, isn't it? Jun 25 06:52:07 mickeyl: One for opie-taskbar and another one for libopie2 Jun 25 06:52:12 SirFred: basically, yes. all the places where transformed is hardcoded, odevice needs to be queried Jun 25 06:52:23 this is libqpe Jun 25 06:52:26 libopie2 Jun 25 06:52:28 opie-taskbar Jun 25 06:52:31 :) Jun 25 06:52:46 mickeyl: Perhaps this is queried from too many places. Jun 25 06:52:57 ram123: sounds like the bootstrap ROM w/o any GUI to me Jun 25 06:53:22 SirFred: yes. while you're at it you could check if it's really necessary to look it up that often Jun 25 06:53:53 mickeyl: I'm not ready for that. Jun 25 06:54:05 mickeyl: It was really hard for me to find the right places to make that last changes. Jun 25 06:54:17 SirFred: sure, don't worry about that nw Jun 25 06:54:17 now Jun 25 06:54:27 * mickeyl greets the awakening kergoth Jun 25 06:54:36 mickeyl: Sorry, but my knowledge of how all this works is really limited. Jun 25 06:54:50 no need to apologize. you're learning pretty fast! Jun 25 06:54:52 mickeyl: I only wanted a w100 driver. (snif) Jun 25 06:54:55 hehe Jun 25 06:55:07 03mickeyl 07 * r1.3616.1.1 10openembedded/packages/busybox/busybox_1.00.bb: busybox: back out the CELF patch since it breaks the return behaviour of 'test'# Jun 25 06:55:11 mickeyl: yah. I don't know if that patch has ever been sent to the busybox people. Jun 25 06:55:13 CoreDump|home: here we are Jun 25 06:55:39 it's only a performance optimisation anyway, so not especially critical. Jun 25 06:55:42 mickeyl: thanks Jun 25 06:56:37 hey Jun 25 06:57:42 i am gettin the following error when i do bitbake bootstrap ? Jun 25 06:57:47 what to do ? Jun 25 06:57:51 http://pastebin.com/303828 Jun 25 06:58:31 hmm Jun 25 06:58:31 configure: error: source directory already configured; run "make distclean" there first Jun 25 06:58:35 i've seen that before Jun 25 06:58:41 does the wiki contain an entry about that? Jun 25 06:58:46 PossibleProblems or so Jun 25 06:58:57 which directory ? Jun 25 06:59:08 the working directory for libtool Jun 25 06:59:16 mickeyl: iirc thats the "running a build inside a path that includes a symlink" one Jun 25 06:59:27 kergoth: ah righto. you and your memory :) Jun 25 06:59:31 heh :) Jun 25 06:59:59 did not get you ? Jun 25 07:00:04 mickeyl: I know that opie-pcmciaapplet is work-in-progress, but is it supposed to be compilable (cvs or 1.2.x tag)? Jun 25 07:00:22 CoreDump|home: yes, it is supposed to be compilable. Jun 25 07:00:26 _cvs Jun 25 07:00:34 heh doesn't for me =) Jun 25 07:00:43 CoreDump|home: pastebin the error for me Jun 25 07:00:52 will do Jun 25 07:02:04 sriram: do you have OE in a path that includes a symlink? If so, don't do that. It confuses libtool Jun 25 07:02:15 as if libtoll wasn't confused enough in the first place Jun 25 07:02:16 heh Jun 25 07:02:32 did not get you ? Jun 25 07:02:37 uhm Jun 25 07:02:45 sorry, i can't rephrase it once more. i'm out of synonyms Jun 25 07:02:54 afternoon Jun 25 07:02:59 hi reenoo_ Jun 25 07:03:09 hey mickeyl Jun 25 07:03:35 mickeyl: http://pastebin.ca/15842 Jun 25 07:04:05 ah Jun 25 07:04:12 opcmciasystem.h not found Jun 25 07:04:16 yeah Jun 25 07:04:16 sounds like libopie2 is too old Jun 25 07:04:24 * CoreDump|home checks Jun 25 07:04:45 how to find out whether it has a simlink or not ? Jun 25 07:05:03 crawl along the path to your directory using ls -l Jun 25 07:10:24 mickeyl: tanks, the applet compiled fine w/ libopie2_cvs. (I'm still using 1.2.0 here) Jun 25 07:10:28 :) Jun 25 07:10:52 mickeyl: would probably be faster to just do a readlink -l (or whatever) to canonicalize the path and compare Jun 25 07:10:56 course that doesnt tell you _where_ the link is Jun 25 07:11:42 mickeyl: Could you suggest a good name for the libopie2 patch ? Jun 25 07:11:55 mickeyl: I'm not good at patch names, as you sure have found. Jun 25 07:12:07 hehe Jun 25 07:12:08 indeed Jun 25 07:12:12 i have taken the freedom to adapt :) Jun 25 07:12:35 mickeyl: Of course. How is it named now? Jun 25 07:13:11 oh just the w100 is now called c7x0-add-w100-accell Jun 25 07:13:12 (or so) Jun 25 07:13:28 honordrivername is actually pretty good Jun 25 07:14:20 mickeyl: And for this one? Perhaps def-qte-driver-support ? Jun 25 07:14:54 fair enough. we will push this upstream asap if it works anyway Jun 25 07:15:33 i am getting the same error ?. and there are no simlinks ? Jun 25 07:15:50 there are no simlinks below my oe directory Jun 25 07:18:01 kergoth: when do we switch to monotone officially? Jun 25 07:18:12 kergoth: I mean it has lua scripting, what else could one want? Jun 25 07:18:25 lol Jun 25 07:18:25 zecke: rotating 3d logo Jun 25 07:18:29 1st of july Jun 25 07:18:29 03CoreDump 07 * r1.3618.1.1 10openembedded/packages/meta/task-bootstrap.bb: task-bootstrap: Add fuser and setserial Jun 25 07:18:30 hehe. how's the testing been going? i've been out of the loop Jun 25 07:19:05 kergoth: I've just imported oe into my db, and have a local tree Jun 25 07:19:08 kergoth: pretty good, the nslu2 people have hit a few snags, but nothing unsolvable Jun 25 07:19:22 kergoth: and the cli syntax is different Jun 25 07:19:30 koen: what should monotone rename do? Jun 25 07:19:36 koen: I've just pushed an org.openembedded.nslu2-linux branch which should solve the nslu2-linux snags Jun 25 07:19:43 zecke: 'bm mv' iirc Jun 25 07:20:00 koen: ok but it did not do it :} Jun 25 07:20:01 ehm, bk Jun 25 07:20:03 * Luke-Jr glares at koen for adding packages w/o DESCRIPTION Jun 25 07:20:20 Luke-Jr: which one? Jun 25 07:20:25 koen: gpe-mini-browser Jun 25 07:20:46 that one doesn't need a description Jun 25 07:20:50 does so =p Jun 25 07:20:53 what is it? Jun 25 07:21:01 a file browser? web browser? Jun 25 07:21:02 a mini-browser for gpe Jun 25 07:21:24 koen: you mean a mini _web_ browser (key part missing) Jun 25 07:22:03 what other browsers are there? Jun 25 07:22:12 bluelightning: that should be kind of obvious from the deps ;) but yeah, there should be a DESCRIPTION Jun 25 07:22:20 you probably mean file*managers* Jun 25 07:22:25 ~lart sdcontrol for beeing an ugly hack Jun 25 07:22:25 * ibot farts in sdcontrol's general direction for beeing an ugly hack Jun 25 07:22:26 gopher browser? Jun 25 07:22:31 reenoo_: not everyone recognises the deps =p Jun 25 07:22:48 koen: FTP browsers =p Jun 25 07:22:52 if it bothers you so much, send a patch Jun 25 07:22:58 otherwise: SHUT THE FUCK UP Jun 25 07:23:05 hey, none of that Jun 25 07:24:09 kergoth: I think monotone is okay, and in one year we might fix bugs ourselves ;) Jun 25 07:24:11 ...anyway Jun 25 07:24:24 * Luke-Jr goes back to growling at KDE's build system for forgetting it needs -lqt-mt Jun 25 07:24:48 Luke-Jr: fix kde's libtool and autotools usage Jun 25 07:25:03 zecke: hmm? that part was mostly working Jun 25 07:26:39 but as soon as I made the native one, it forgot about Qt Jun 25 07:32:44 mickeyl: Just a question about this code. Jun 25 07:33:14 mickeyl: My first impulse was to make a member for ODevice.h with this prototype: 'virtual const QString &qteDriver() const' Jun 25 07:34:02 mickeyl: But I've seen that this is not the "fashion" in this class. All the methods use to return QString and not a const QString&. Is there any reason? Jun 25 07:34:36 mickeyl: I always thought that returning const references to the members is a lighter way to do this kind of things. Jun 25 07:34:52 SirFred: indeed it would be slightly lighter Jun 25 07:35:20 SirFred: but we follow Qts style for returning strings (and they are shared) Jun 25 07:35:24 zecke: Yes, but my method seems strange, totally rounded with heavy QString returners. Jun 25 07:35:36 SirFred: so follow the other methods Jun 25 07:35:47 SirFred: note with virtual you will break binary compatibility Jun 25 07:35:56 mickeyl: should libopie2 remain bc? Jun 25 07:36:17 zecke: Am I not breaking binary compatibility anyway, adding new methods to ODevice ? Jun 25 07:36:43 zecke: You mean that QString has an internal method for not duplicating, something like count references or so? Jun 25 07:36:51 s/count references/reference count/ Jun 25 07:36:56 SirFred: it is shared Jun 25 07:37:21 SirFred: I mix up impilicit and explicit but it is shared ;) Jun 25 07:38:04 zecke: Well the method could be not virtual anyway. Jun 25 07:38:05 zecke: I don't think it's that much of a value to stay BC at this point. libopie2 has always been in flux and is far from being finished. Adding everything through virtual hooks at this point of time would just be annoying. I agree that it's kind of backwards for a stable tree, but I really don't think it'll be worth it. It's not like many people are using our API except us anyway :) Jun 25 07:38:14 SirFred: http://developer.kde.org/documentation/other/binarycompatibility.html Jun 25 07:38:27 SirFred: this is why we have that damn ugly virtual hook Jun 25 07:38:58 mickeyl: I wouldn't be too confident on the part of the last statement Jun 25 07:38:59 zecke: I'm going to read it. Thanks! Jun 25 07:39:40 SirFred: but yes it will be virtual in one way or another Jun 25 07:40:05 SirFred: just make it work, not look too alienated from the other methods Jun 25 07:40:17 and then we can decide on BC Jun 25 07:40:20 zecke: Well, it's going to work anyway. Jun 25 07:40:49 zecke: Making it virtual was only useful to reimplement it in subclasses. It's not necessary now. Jun 25 07:41:09 zecke: i know who you are thinking about, however there are really some more changes in the pipeline, i.e. I have seen that I need to change QStringList to QString in OPcmciaManager::productIdentity() since all clients using that code just do join().stripWhiteSpace() anyway. Noone is using the entries directly hence there's no point returning them. Jun 25 07:41:12 SirFred: for such an attribute you can always have a protected setQteDriver Jun 25 07:41:43 mickeyl: hehe the return value is not mangled into the signature Jun 25 07:41:47 mickeyl: one can change it :} Jun 25 07:42:03 oh Jun 25 07:42:06 not that I recommend it Jun 25 07:42:07 fine Jun 25 07:43:14 zecke: That's one of the reasons one can not have two methods with the same name, same args and different return value, isn't it? Jun 25 07:43:56 SirFred: right, but that is demanded by C++ Jun 25 07:44:03 no polymorphism on return values Jun 25 07:46:49 Interesting reading. Jun 25 07:47:12 So this d-pointer role is what is playing ODeviceData in ODevice. Jun 25 07:47:43 And so, adding a m_qteDriver to ODeviceData, and just a non virtual function to ODevice won't break binary compatibility. Jun 25 07:47:53 getting the same error ; no simlinks found below /stuff directory ? Jun 25 07:47:58 what could be wrong ? Jun 25 07:48:08 zecke, mickeyl. Thanks a lot Jun 25 07:50:57 np Jun 25 07:51:50 please refer posting: http://pastebin.com/303828 Jun 25 07:52:28 no sim links found and i have make distclean ; make clean ; but getting the same error Jun 25 07:53:57 well, there's certainly something suspicious going on there. Are you certain that /usr/tmp and /var/tmp are not symlinked to each other? Jun 25 07:58:52 omg Jun 25 07:58:59 KDE's configure is stupid Jun 25 07:59:13 It tests for one thing and uses the answer in config.h for something else Jun 25 07:59:46 16:24 < zecke> Luke-Jr: fix kde's libtool and autotools usage Jun 25 08:00:24 koen: yes, but I thought that was just for cross-compiling... Jun 25 08:00:30 koen: this is a problem with -native Jun 25 08:02:15 1 sec Jun 25 08:02:49 yes it is ? Jun 25 08:02:56 what do i do now ?. Jun 25 08:03:16 i thought sim link is only for /stuff and below ?. Jun 25 08:03:55 where did you get that idea? Jun 25 08:04:08 you mustn't have a symlink anywhere in the path to your TMPDIR. Jun 25 08:04:19 wb bluelightning_ Jun 25 08:04:36 as for what you do now, my recommendation would be to change TMPDIR to a path that does not include any symlinks Jun 25 08:05:11 thats the only reqt or at anyother place sim link should not exit. Jun 25 08:05:30 I think TMPDIR is the only one that matters. Jun 25 08:05:37 symlinks in the path to the sources should be harmless. Jun 25 08:06:33 will changing the TMPDIR in local.conf will do or anywhere else i have to make changes ? Jun 25 08:06:39 please confirm ? Jun 25 08:10:09 ok i want to build cross tools ; x windows and GPE what commands should i give to bitbake Jun 25 08:10:33 bitbake gpe-image Jun 25 08:10:42 1 sec Jun 25 08:11:00 i am building from scracth. no cross compiler, libraries etc Jun 25 08:11:08 what command should i give first. Jun 25 08:11:45 7 projects for handhelds.org from Google Summer of Code :) Jun 25 08:12:00 sriram: he just told you. Jun 25 08:12:12 sriram: oe builds itself a crosscompiler when it needs a crosscompiler. Jun 25 08:12:20 ok Jun 25 08:12:40 http://groups-beta.google.com/group/summer-discuss/browse_frm/thread/8b5b85e38fc8efbc/62024b5ec92784f7#62024b5ec92784f7 Jun 25 08:12:47 suppose i want gpe does it build xserver, gtk and gpe Jun 25 08:12:53 yes Jun 25 08:12:56 or only gtk and gpe Jun 25 08:13:05 it can't build gtk w/o X Jun 25 08:13:10 it will build an image with everything needed Jun 25 08:13:17 Luke-Jr: sure it can Jun 25 08:13:22 Luke-Jr: gtkfb Jun 25 08:13:32 koen: o.o Jun 25 08:13:49 gtk has framebuffer fetishists too, just like qt Jun 25 08:14:01 * Luke-Jr pretends it doesn't exist Jun 25 08:14:17 and later if i want to add something new ; will it start from scratch or add to the existing stuff. Jun 25 08:14:38 sriram: it will use older material Jun 25 08:14:41 unless you delete it Jun 25 08:14:59 where are these stored ? Jun 25 08:15:07 tmp dir Jun 25 08:15:53 and for the cross compiler and cross libraries also is it stored in TMPDIR Jun 25 08:16:09 yes Jun 25 08:16:14 <_chronic> sriram: have you read http://oe.handhelds.org/cgi-bin/moin.cgi/GettingStarted ?? Jun 25 08:16:19 yes Jun 25 08:16:47 <_chronic> sriram: and http://oe.handhelds.org/cgi-bin/moin.cgi/DirectoryStructure Jun 25 08:17:30 yes sir. i have read it and trying to understand what ever is written Jun 25 08:18:44 tmp/ - $TMPDIR Jun 25 08:18:44 * Jun 25 08:18:44 All of OE output ends up here. Jun 25 08:18:45 just moved oetmp to newdir, changed local.conf to point to newdir and did bitbake bootstrap-image and still getting the same error. Jun 25 08:18:47 heh Jun 25 08:19:18 changed OEROOT? Jun 25 08:19:24 err Jun 25 08:19:28 no Jun 25 08:19:32 BBPATH* Jun 25 08:20:16 BBPATH does not point to TMPDIR Jun 25 08:20:32 no Jun 25 08:20:41 but to the dir w/ conf/local.conf Jun 25 08:20:52 so only placees i must change is TMPDIR Jun 25 08:21:01 am i right ? Jun 25 08:21:51 if you're moving only TMPDIR... Jun 25 08:21:59 yes Jun 25 08:22:02 do we know what the 7 Summer of Code projects are yet? Jun 25 08:22:36 do we have a log of this chat anywhere on the net ? Jun 25 08:22:36 see #handhelds.org :) Jun 25 08:22:43 logs Jun 25 08:22:48 ~logs Jun 25 08:22:48 chat logs ? Jun 25 08:22:48 apt/ibot/jbot/purl all log to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz, or updated "nightly" Jun 25 08:25:01 thanks Jun 25 08:25:17 mickeyl: I've just sent you two more patches. Jun 25 08:25:24 i was making the log myself for future reference. Jun 25 08:25:32 mickeyl: Tested them and seems to work. Jun 25 08:27:00 do you have any idea how much disk space it takes for bitbake gpe_image from scratch. Jun 25 08:28:31 5GB Jun 25 08:29:39 suppose i have build gpe_image and want to change defconf of kernel. does bitbake does all of them allover again or it just compiles the kernel Jun 25 08:30:08 sriram: build the kernel, build the image afterwards Jun 25 08:30:12 just the kernel Jun 25 08:32:06 <_chronic> sriram: one of the points of building and defining metadata dependencies is to minimise rebuilding, ie only update what you need. Jun 25 08:33:04 building process hangs at http://pastebin.com/303865 Jun 25 08:33:11 please refer http://pastebin.com/303865 Jun 25 08:35:29 it seems to be checking out a cvs tree Jun 25 08:35:59 for the last 20 mins Jun 25 08:37:28 are you guys available tommorrow here Jun 25 08:37:42 just in case i need ur help Jun 25 08:39:40 sure Jun 25 08:39:45 03koen 07 * r1.3616.2.1 10openembedded/ (6 files in 4 dirs): Jun 25 08:39:45 All gnome-vfs versions: add plugin-file and mime-date to RRECCOMMENDS Jun 25 08:39:45 shared-mime-info_0.16.bb: new file Jun 25 08:40:08 another meta-bug squashed Jun 25 08:40:16 ~lart strict mime checking Jun 25 08:40:16 * ibot tries to shut strict mime checking up Jun 25 08:40:16 Ciao all Jun 25 08:40:21 hey Pigi Jun 25 08:40:29 thanks koen. without ur and others over here i would have struggled. thanks a million... Jun 25 08:42:29 see you tommorrow. Jun 25 08:42:35 take care Jun 25 08:52:18 koen: could you add my key to your monotone db? Jun 25 08:52:28 koen: and can one mess with the trees? Jun 25 08:52:42 zecke: yes, and yes Jun 25 08:53:14 zecke: use zecke@openembedded.org as an ID Jun 25 08:53:20 otherwise the regexp will fail Jun 25 08:53:42 zecke: mail it to koen@handhelds.org Jun 25 08:54:22 koen: oh Jun 25 08:55:31 koen: how do I export my key? Jun 25 08:55:45 ah found it Jun 25 08:55:46 03koen 07 * r1.3622 10openembedded/packages/evince/ (evince_0.3.2.bb evince/no-doc.patch): Add evince, a poppler based pdf/document viewer Jun 25 08:56:07 wth? Jun 25 08:56:10 http://oe-devel.bkbits.net:8080/openembedded/stats?nav=index.html Jun 25 08:56:31 I've finally beaten mickeyl Jun 25 08:57:05 (I am however 3k cset behind in the overall score) Jun 25 08:58:44 hey ar_ Jun 25 08:58:56 hi zecke Jun 25 09:00:06 zecke: key is in Jun 25 09:00:51 thank Jun 25 09:03:49 he obi Jun 25 09:17:16 mickeyl: time for fed cup? Jun 25 09:28:32 koen: how would I create my local clone? Jun 25 10:09:17 'lo all Jun 25 10:11:40 hi Jun 25 10:39:11 hi Jun 25 10:39:59 * Pigi is away: I'm busy Jun 25 11:12:57 hey Jun 25 11:17:24 hmm Jun 25 11:17:33 I think I found the cause of the suspend-on-right-arrow problem on the SL-5500 Jun 25 11:17:50 and according to this, it should also suspend on center button and on menu button Jun 25 11:25:29 marcan: negative, I just tried the center and menu button. It didn't suspend Jun 25 11:25:45 mine does generate the events Jun 25 11:26:04 does right arrow suspend anyway? Jun 25 11:26:11 ok, menu button does suspend :) Jun 25 11:26:16 ok :) Jun 25 11:26:22 got the wrong button the first time heh Jun 25 11:27:17 soo... the LoCoMo chip is treating all keys on that row as suspend requests Jun 25 11:27:21 and setting the GPIO pin Jun 25 11:27:29 I smell misconfiguration somewhere Jun 25 11:32:54 someone can send me a local.conf file for a poodle ? Jun 25 11:35:30 has to do with apm Jun 25 11:35:42 b/c with apm turned off, it doenst suspect Jun 25 11:35:53 of course, that could just be a result, not the cause Jun 25 11:35:54 the apm problem is the crashing on suspends Jun 25 11:36:17 the problem of right arrow suspending is either: Jun 25 11:36:30 the LoCoMo is misconfigured and is sending ON requests when it shouldn't Jun 25 11:36:30 or Jun 25 11:36:34 the question becomes, why just the right arrow? Jun 25 11:36:44 that *is* what it should do, but we should filter further Jun 25 11:36:47 so one tends to think maybe keyboard configuration? Jun 25 11:36:48 right arrow, menu button, and center button Jun 25 11:36:56 are all on the same physical keyboard row connection Jun 25 11:37:04 (and cancel button) Jun 25 11:37:16 menu also suspends, and cancel also does Jun 25 11:37:18 unfortunately, no one seems to be interested Jun 25 11:37:19 (for me at least) Jun 25 11:37:25 *I* am interested Jun 25 11:37:31 I've just flashed oz 1.3.1 kernel Jun 25 11:37:38 and I'll check if the GPIO is supposed to do that or no Jun 25 11:45:11 marcan: your research on this subject is most appreciated :) Jun 25 11:45:25 thanks :) Jun 25 11:45:32 I've just flashed the kernel from OZ 3.5.1 Jun 25 11:46:01 aside the fact that usbnet is broken (maybe module incompatibility or something, but I honestly don't care about that), it still resumes with right arrow Jun 25 11:46:06 is it supposed to do that with OZ 3.5.1? Jun 25 11:46:15 it might be a userland change causing the problem Jun 25 11:46:27 s/resumes/suspends Jun 25 11:49:12 mickeyl: are you there? Jun 25 11:54:37 IIRC 3.5.1 was fine. People only started bitching about it since 3.5.3. Jun 25 11:55:22 marcan: I'll flash a stock IZ 3.5.1 / 2 image and tell you the results Jun 25 11:55:41 thanks Jun 25 11:55:52 I've flashed the 3.5.1 kernel and the right arrow thing is still present Jun 25 11:56:06 so it might be userland messing it up somehow Jun 25 11:56:20 (I've still got current userland) Jun 25 11:57:02 there wasn't much work on collies kernel for some time now. Userland seems indeed be the most likely cause Jun 25 11:58:53 where is the kernel dir supposed to be in the oe environment ? Jun 25 12:00:29 Pigi: you mean where the kernel image is unpacked and built? Jun 25 12:01:05 marcan, not exactly. I need to pass to alsa-driver the kernel location so the autoconf would read the config file. Jun 25 12:01:59 tmp/work//linux/ Jun 25 12:02:07 ${STAGING_KERNEL_DIR} Jun 25 12:02:38 ok Jun 25 12:02:46 see other moudles (hostap, cfu1u) for details Jun 25 12:02:55 that's a good idea Jun 25 12:06:38 * CoreDump|home is flashing OZ 3.5.1 w/ a 56-8 kernel Jun 25 12:09:02 is 2.6 on collie/poodle progressing? Jun 25 12:10:09 * CoreDump|home wonders what the point of a 2.6 collie kernel w/o SD support is Jun 25 12:10:42 write an sd driver? Jun 25 12:10:49 * marcan wonders what the point of a binary sd driver is Jun 25 12:10:50 didn't kergoth have the specs? Jun 25 12:11:14 * marcan wonders what the point of a closed spec memory standard is Jun 25 12:11:44 marcan: the spec is open Jun 25 12:11:53 it wasn't not too long ago Jun 25 12:12:03 not for collies retarded controller AFAIK Jun 25 12:12:08 about a year Jun 25 12:12:23 SD spec != chip spec Jun 25 12:12:57 but alsa 1.0.8 doesn't have support for uda1380... Jun 25 12:13:46 try asking jamey or pb_ which driver to use Jun 25 12:13:59 I will Jun 25 12:14:08 pb_ you around ? Jun 25 12:15:18 Anyone know if there's util capable of resizing partitions in oe? Jun 25 12:15:19 marcan: OZ 3.5.1: "Cancel" suspends and resumes perfectly. Neither "menu" nor "right-arrow" suspend the device. Jun 25 12:15:33 can you test something out for me? Jun 25 12:15:36 try modprobe registers Jun 25 12:15:43 sure Jun 25 12:16:57 loaded Jun 25 12:17:00 now Jun 25 12:17:21 while true ; do cat /proc/cpu/registers/GPLR ; done Jun 25 12:17:24 and check the last digit Jun 25 12:17:37 it should be 1 most of the time Jun 25 12:19:08 umm, it's not Jun 25 12:19:14 what is it? Jun 25 12:19:26 it switches between "3" "B" and "F" Jun 25 12:19:35 you have AC connected, have you? Jun 25 12:19:37 this is over SSH / WLAN Jun 25 12:19:39 yes# Jun 25 12:19:42 then 3 is fine Jun 25 12:19:53 ok, press cancel a short time Jun 25 12:19:54 and watch Jun 25 12:19:58 it should switch to 2 Jun 25 12:20:25 correct Jun 25 12:20:32 now try right arrow and menu Jun 25 12:21:05 same for both Jun 25 12:21:13 aha Jun 25 12:21:22 ok, ctrl+c that Jun 25 12:21:30 yup Jun 25 12:21:34 now, surely right arrow doesn't reset, does it? Jun 25 12:21:49 nope, not in Opie and not in the VT Jun 25 12:21:57 ok Jun 25 12:22:00 try killall apmd Jun 25 12:22:09 and try to suspend with right arrow, then cancel Jun 25 12:22:18 try with and without Opie running Jun 25 12:23:29 hah Jun 25 12:23:44 ok, Cancel suspends in Opie and VT Jun 25 12:24:03 right-arrow does _not_ suspend in Opie but it does in VT Jun 25 12:24:11 aha Jun 25 12:24:21 I think I know what we're leading up to Jun 25 12:24:58 yeah, opie did some magic in the past to work around the right-arrow Jun 25 12:25:10 OK, I got it Jun 25 12:25:20 for starters, GPIO<0> is NOT on-button Jun 25 12:25:26 it is any button in that row Jun 25 12:25:34 including right arrow and all the others Jun 25 12:25:39 so using that for suspending is useless Jun 25 12:25:51 I have no idea why sharp wired it that way, but it is useless Jun 25 12:25:58 hmm Jun 25 12:26:06 at least without checking first for the Cancel press from they keyboard driver Jun 25 12:27:13 there *is* a userland way of disabling the Cancel suspend Jun 25 12:27:15 Pigi: at your service Jun 25 12:27:23 at least using the DAEMON APM mode Jun 25 12:27:31 but it is still dumb using it in the first place Jun 25 12:27:50 so I'm patching that to at least check for Cancel from the keyboard driver Jun 25 12:29:43 pb_ I would like to try compile an alsa driver for h3900 but in 1.0.8 there isn't an uda1380 Jun 25 12:29:54 Pigi: you need to use the version from handhelds.org cvs Jun 25 12:30:11 0.9. 6 ? Jun 25 12:30:43 dunno. if that's the latest in there, yes. Jun 25 12:30:50 ok thx Jun 25 12:31:13 you can probably copy the uda1380 driver from there into a newer alsa tree, if you want Jun 25 12:32:43 It could be, but actually alsa 1.0.8 doesn't even finish a configure with our recent kernel config :) Jun 25 12:37:43 * CoreDump|home flashes back to 3.5.4-pre Jun 25 12:38:23 Is here any simpad sl4 users? I need some help for flashing new image. Jun 25 12:42:46 god damn checkversions Jun 25 12:44:47 Pigi: heh Jun 25 12:52:12 bbl Jun 25 12:52:18 * Pigi is away: I'm busy Jun 25 12:53:29 mickeyl: what do you think about editing checkversions to only check for YYYYMM, not YYYYMMDD for snapshot releases? Or drop the check for snapshots alltogether Jun 25 12:54:00 having checkversions bark because of DD sucks big time :\ Jun 25 12:54:31 * marcan is compiling the fixed kernel Jun 25 12:54:40 CoreDump|home: the right-button and related bug should be gone now Jun 25 12:54:43 if it compiles properly Jun 25 12:54:52 it now checks the keyboard matrix for Cancel as well as the GPIO input Jun 25 12:54:53 great! Jun 25 12:55:28 left is the suspend hanging Jun 25 12:55:38 but that definitely is some kernel/apmd breaking Jun 25 12:55:41 probably userland Jun 25 12:57:11 * marcan is flashing new kernel... Jun 25 12:58:26 marcan: probably userland breaking kernelspace Jun 25 12:58:36 it is embeddix code after all... Jun 25 12:58:36 yep Jun 25 12:58:41 heh Jun 25 12:58:51 well at least the incredibly dumb right arrow bug is gone Jun 25 12:59:05 I should replace all kernel defines of that GPIO from ON_KEY to ON_ROW Jun 25 12:59:12 because it ISN'T only the ON key Jun 25 12:59:25 marcan: apmd hasn't been touched for quite a while. The newest change was 10 weeks ago: Jun 25 12:59:34 http://oe-devel.bkbits.net:8080/openembedded/diffs/packages/apmd/apmd-3.2.2/apmwrapper@1.5?nav=index.html|src/|src/packages|src/packages/apmd|src/packages/apmd/apmd-3.2.2|hist/packages/apmd/apmd-3.2.2/apmwrapper Jun 25 13:00:10 !test Jun 25 13:00:11 Failed! Jun 25 13:00:18 !tiny http://oe-devel.bkbits.net:8080/openembedded/diffs/packages/apmd/apmd-3.2.2/apmwrapper@1.5?nav=index.html|src/|src/packages|src/packages/apmd|src/packages/apmd/apmd-3.2.2|hist/packages/apmd/apmd-3.2.2/apmwrapper Jun 25 13:00:20 CoreDump|home: http://tinyurl.com/9tqd7 Jun 25 13:00:55 hmm Jun 25 13:04:59 * marcan recompiles after fixing stupid mistake on my part Jun 25 13:10:18 Perfect fix :) Jun 25 13:10:27 now only Cancel does it Jun 25 13:11:14 ~praise marcan Jun 25 13:11:14 All hail marcan! Jun 25 13:11:41 lol Jun 25 13:13:22 'ere Jun 25 13:13:27 who do I send the patch to? :P Jun 25 13:13:56 oe@hentges.net, I'll push it Jun 25 13:14:39 do I just send the kernel patch, or do I put it into the OE tree and change the bb file to apply it? Jun 25 13:15:44 * Pigi is back (gone 00:23:26) Jun 25 13:16:08 How can I convince oe to execute a script instead af the configure ? Jun 25 13:16:31 CoreDump|home: ? Jun 25 13:17:23 Pigi: maybe write a do_configure() in your BB Jun 25 13:17:31 < marcan> do I just send the kernel patch, or do I put it into the OE tree and change the bb file to apply it? Jun 25 13:17:31 marcan: the patch will do Jun 25 13:17:35 ok Jun 25 13:17:49 i did it, but he ( oe ) doesn't like it very much :) Jun 25 13:17:53 trying again Jun 25 13:20:12 oh, wtf, my DNS servers are down Jun 25 13:20:20 heh Jun 25 13:21:01 if it isn't to big, you can put it into pastebin Jun 25 13:21:30 it's tiny Jun 25 13:21:36 yeah, pastebin'll do Jun 25 13:23:01 * CoreDump|home hugs sed Jun 25 13:25:34 CoreDump|home: http://pastebin.ca/15852 Jun 25 13:38:34 marcan: if you cant find anyone for that, kergoth or lardman should be able to help Jun 25 13:39:05 marcan: I wouldnt mind if you made a fix that people could simply download Jun 25 13:39:37 marcan: this is as far as I got http://www.oesf.org/index.php?title=Opie_3.5.3_on_the_5500http://www.oesf.org/index.php?title=Opie_3.5.3_on_the_5500 Jun 25 13:40:22 marcan: any changes are welcome Jun 25 13:40:31 marcan: attach it to the bugzilla please Jun 25 13:41:07 *sigh* Jun 25 13:41:22 note: gpe doesnt have the darn right arrow buggaboo Jun 25 13:41:24 even with big "OPIE 1.2" screens people still don't get the point Jun 25 13:41:32 OZ=distro, opie=env Jun 25 13:42:02 then again, the same people think ROMs are RW Jun 25 13:42:14 * koen|tv gives up on the dumbasses@oesf Jun 25 13:42:58 subdue: gpe must override the kernel for that Jun 25 13:43:05 but the bug *is* in the kernel Jun 25 13:43:19 even in OZ 3.5.1 Jun 25 13:43:20 koen|tv: you mean the forums? Jun 25 13:43:59 that wiki is also a nice proof of the ignorance Jun 25 13:44:09 or some timewarp Jun 25 13:44:22 3 years into the future: opie 3.5.3 Jun 25 13:44:37 marcan: if you fix the darn oz issues, then I'll be able to try it again on my 5500! Jun 25 13:44:50 koen|tv: which one? Jun 25 13:44:59 the one you just pasted Jun 25 13:45:11 oh, that's one I created, trying to use it. it's my fault Jun 25 13:45:27 it was my experiences, sorry Jun 25 13:45:30 the right-arrow stuff is fixed. I'll try to look at the other (suspend hanging) bugs Jun 25 13:45:32 feel free to fix it Jun 25 13:45:39 marcan: thanks Jun 25 13:45:53 koen|tv: I'm new to opie, and found myself digging for info Jun 25 13:46:06 hence my hoping that others didnt have to go through the same problems I did Jun 25 13:46:22 ehm, try opening http://opie.handhelds.org Jun 25 13:46:29 top news item Jun 25 13:46:53 * koen|tv still pissed at oesf whiny bitches refusing to report bugs into bugzilla Jun 25 13:47:14 if you tell me what I need to do, I'll put it in there Jun 25 13:47:22 subdue: not you Jun 25 13:47:25 oh. ok Jun 25 13:48:09 a lot of the folks do something like "I reported the bug 3 *days* ago on this forum, and it still isn't fixed, you guys suck!" Jun 25 13:48:21 pb_ still around ? Jun 25 13:48:28 ah ... stupidity, right Jun 25 13:48:52 I've been fighting that on the forums Jun 25 13:48:56 that's why we have this: Jun 25 13:48:58 ~users Jun 25 13:48:58 so this is what we get for sharing our unpaid volunteer work with you? complaints and accuses? well done, this clearly supports our motivation to continue working on open source projects. Jun 25 13:49:34 hmm Jun 25 13:49:43 don't try building a dir while rming it at the same time Jun 25 13:49:45 =p Jun 25 13:49:46 ~forums Jun 25 13:51:30 ibot, forum is where people goes complaining about the work volunteers do for their plaisure Jun 25 13:51:31 ...but forum is already something else... Jun 25 13:51:38 ~forum Jun 25 13:51:38 Another PHP3/Mysql forum with some nifty features. URL: http://www.itcenter.ru/eng/forum/ Jun 25 13:51:49 ~lart pastebin for messing with whitespace Jun 25 13:51:49 * ibot readies the nuke launcher and fires some rounds at pastebin for messing with whitespace Jun 25 13:51:53 ibot, forums is where people goes complaining about the work volunteers do for their plaisure Jun 25 13:51:54 okay, Pigi Jun 25 13:52:00 ~forums Jun 25 13:52:01 rumour has it, forums is where people goes complaining about the work volunteers do for their plaisure Jun 25 13:52:19 Pigi: any luck with alsa yet? Jun 25 13:52:42 almost. I need some input from pb, but I have passed the configure phase Jun 25 13:52:58 cool Jun 25 13:53:49 CoreDump|home: isn't there a "raw download" link? Jun 25 13:53:50 but I'm working on 0.9.6 from handhelds.... wishing this wouldn't need also lib and tools from 0.9.6 :) Jun 25 13:53:53 anyway, I added it to bugzilla Jun 25 13:53:59 http://bugs.openembedded.org/show_bug.cgi?id=54 Jun 25 13:54:20 Pigi: according to pb_ a few days ago, you shouldn't need the utils and libs Jun 25 13:54:26 only drivers with oss emu Jun 25 13:54:36 marcan: indeed, didn't know about that Jun 25 13:55:30 cool, bug 74 was marked as a duplicate Jun 25 13:55:57 I marked it :P Jun 25 13:56:04 * koen|tv likes duplicates: easily squashed Jun 25 13:56:08 ~hail marcan Jun 25 13:56:08 * ibot bows down to marcan and chants, "I'M NOT WORTHY!!" Jun 25 13:56:17 lol Jun 25 13:56:36 (now if you can fix the gui ipkg issues) Jun 25 13:56:59 heh Jun 25 13:57:04 don't use the gui ;) Jun 25 13:57:11 lol Jun 25 13:58:50 if you want extra features in ipkg, talk to pigi Jun 25 13:59:01 pigi loves adding features, right pigi? Jun 25 13:59:19 haha, just want it to work is all Jun 25 13:59:30 koen|tv, ehehe Jun 25 13:59:30 probably just needs more msgs Jun 25 13:59:55 I do prefere adding feature, than fixing bugs :) Jun 25 14:04:19 03CoreDump 07 * r1.3623 10openembedded/packages/linux/ (2 files in 2 dirs): Collie: Don't suspend on right-arrow. Courtesy of marcan. Thanks Jun 25 14:06:07 ah, it seems there is a chance to liberate opie from the root-only dark-ages Jun 25 14:06:11 ~hail google SoC Jun 25 14:06:12 * ibot bows down to google SoC and chants, "I'M NOT WORTHY!!" Jun 25 14:06:41 ...google-onna-chip? Jun 25 14:06:54 oh, SoC = Summer of Code. Gotcha. Jun 25 14:08:44 * CoreDump|home marked bug #54 as fixed Jun 25 14:09:08 CoreDump|home: not Jun 25 14:09:19 this fixes the right arrow b0rking Jun 25 14:09:24 but not the suspend /resume failuresd Jun 25 14:09:30 (better than not being fixed!) Jun 25 14:09:41 oh he notes resume, too? Jun 25 14:09:48 sorry Jun 25 14:10:21 he notes everything Jun 25 14:10:28 apm was broken quite badly Jun 25 14:16:44 koen|tv: ? Jun 25 14:20:21 SoC should be Summer onna Chip. Jun 25 14:20:27 There's no summer here yet :( Jun 25 14:28:39 CoreDump|home: can anyone commit to OE or do you need a developer account? Jun 25 14:29:58 you need a dev account Jun 25 14:32:41 koen|tv, do you think we do need sequencers ? Jun 25 14:33:00 me personally, no Jun 25 14:33:09 but the midi freaks might disagree Jun 25 14:33:40 if the midi freaks disagree, they will eventually provide a patch, right ? Jun 25 14:33:40 Pigi: disable it and wait for someone to complain Jun 25 14:33:45 heh Jun 25 14:33:48 MIDI is nice, in certain cases Jun 25 14:34:02 wasn't aware OE had anything supporting it.. Jun 25 14:34:18 Luke-Jr, are you volunteer to port alsa ? :) Jun 25 14:34:26 oooh Jun 25 14:34:28 hardware midi... eww Jun 25 14:34:42 Tell MIDI freaks to use timidity++ Jun 25 14:34:43 :) Jun 25 14:34:48 :) Jun 25 14:35:24 my view on hardware MIDI is that it is at best useless =p Jun 25 14:35:50 Does any OE-supported hardware even have MIDI stuff? Jun 25 14:42:07 I guess you can fit MIDI to anything with PCI. so, yeah, in theory. Jun 25 14:42:56 pb_ do you have any idea why pxa-i2s.h is not in staging kernel area ? Jun 25 14:43:13 Pigi: uh, no. Is that file part of the kernel? Jun 25 14:43:35 well, it seems Jun 25 14:43:42 handhelds-pxa/linux/kernel/drivers/sound/pxa-i2s.h Jun 25 14:43:54 ah, I see, it's in drivers/sound Jun 25 14:44:01 I think that only headers from include/ are staged. Jun 25 14:44:02 right Jun 25 14:44:16 that's a problem for alsa :) Jun 25 14:44:33 okay Jun 25 14:44:42 well, you could take a copy of the header and patch it into the alsa tree. Jun 25 14:44:55 or, alternatively, you could teach the kernel .bb file to copy it into staging. Jun 25 14:45:22 I have thougth that too, but I'm scared about consequences if that file change. Better the second idea Jun 25 14:48:24 Realistically, that file isn't going to change in the 2.4 tree. But, yeah, the second plan would be better. Jun 25 14:48:53 ok. I will see if I can get the drivers compiled, then I will think at it Jun 25 14:49:00 righto Jun 25 14:51:32 can haret dump registers as well? Jun 25 14:51:54 what sort of registers? Jun 25 14:52:12 PXA builtins (memory mapped) Jun 25 14:52:29 yeah, it can dump any memory mapped region Jun 25 14:53:16 ah right Jun 25 14:59:24 Pigi: you can enable virtual sequencers. Useful to have timidity++ as a server and have apps work with /dev/sequencer Jun 25 14:59:54 marcan, actually I will try to give voice to pxa, then we will see :) Jun 25 15:02:03 hmm.. D you think OE would be a good buildsystem for putting linux inside a DSL router? Jun 25 15:02:12 (ARM9 CPU, no MMU, uClinux) Jun 25 15:04:55 I think there are OE distributions for routers from netgear and 4G-systems Jun 25 15:05:32 interesting... Jun 25 15:05:39 I'm trying to port linux to my DSL router Jun 25 15:06:12 the kernel works, and most apps work, but the buildsystem (if it can be called that) supplied with the standard devkit for the CPU is crappy Jun 25 15:06:14 and everything is outdated Jun 25 15:06:47 it's too late for me now. Jun 25 15:06:50 night all Jun 25 15:07:17 marcan: possible, nslu, openwrt use OE to some degree Jun 25 15:07:37 aha Jun 25 15:08:08 I mostly need a system that properly builds apps for the target CPU (MMUless FPUless ARM9) Jun 25 15:08:26 I've tried the uClibc buildroot and similar, but I couldn't get things to work Jun 25 15:08:36 do you know if OE supports FLAT style binaries? Jun 25 15:25:02 marcan: OE is "just" a collection of meta-data. You can do almost everything with it if you dive deep enough into its core. It is very flexible. Jun 25 15:26:25 say, for example, that I create a configuration for the characteristics of the hardware Jun 25 15:26:43 it should then be fairly easy to build the basic busybox/core apps, shouldn't it? Jun 25 15:27:10 yeah Jun 25 15:27:47 cool :) Jun 25 15:27:47 once the machine configuration is done, building apps is trivial. Jun 25 15:29:04 marcan: what sort of DSL router do you plan to add? Jun 25 15:29:47 It's a Samsung S3C2510A01 Jun 25 15:29:55 ARM9TDMI, no MMU, no FPU Jun 25 15:30:15 it's already working with uClinux, but the userland buildsystem that came with the dev kit is useless and outdated Jun 25 15:30:48 currently we are using a prebuilt FS image for testing, but we need to have a working buildsystem Jun 25 15:31:33 The router itself is a 3com 3CRADSL72, identical to the SMC7804WBRA, and similar to lots of other routers Jun 25 15:36:16 'night all Jun 25 15:36:17 'night reenoo_ Jun 25 15:36:23 'night all Jun 25 17:28:18 hey umm... would any of y'all happen to have recent gpsd packages for arm? Jun 25 17:40:25 wrong channel Jun 25 17:40:30 oh Jun 25 17:40:37 this is a developer channel Jun 25 17:42:20 bbshell's reparse doesn't seem to be quite thorough. Jun 25 17:42:42 although I should probably do svn up before saying anything about bbshell, because mickeyl has snuck in fixes on me in the past Jun 25 18:50:16 wtf. this thing reboots when I do 'ipkg install.' But only when the wifi card is plugged in. Jun 25 19:00:21 keturn: If it hurts, don't do it :-P Jun 25 19:10:00 nail: ...but... the wifi card *wants* to be plugged in. Jun 25 19:14:47 Have you asked the wifi card, or are you just assuming that? wifi cards are sensitive buggers. Take heed of the phase of the moon before plugging it in ;-) Jun 25 19:15:00 (Other than that, I don't think I can help you much) Jun 25 19:19:34 keturn: What device? Jun 25 19:32:49 xxs1500 with a wl1100c wifi card Jun 25 20:08:37 night Jun 25 23:21:12 wow Jun 25 23:21:18 * jumpkick bows to marcan Jun 25 23:21:34 awesome stuff Jun 25 23:41:33 mickeyl: ping? **** ENDING LOGGING AT Sat Jun 25 23:59:56 2005