**** BEGIN LOGGING AT Wed Mar 15 10:59:56 2006 Mar 15 11:01:36 hmmh Mar 15 11:01:53 still having problems with ftp.handhelds.org Mar 15 11:03:38 send email to admin@handhelds.org Mar 15 11:04:51 well, if it runs gpe, maybe it can run opie too... :) Mar 15 11:05:18 depending on the price, might be a very interesting toy Mar 15 11:05:38 I'll send them a real professional one. "lol help plz teh ftp has no /pub" Mar 15 11:12:03 morning all Mar 15 11:12:13 hi Liam Mar 15 11:14:14 hey lrg Mar 15 11:15:33 morning all Mar 15 11:15:54 hi Dirk Mar 15 11:16:02 hey Dirk, koen, hrw|work Mar 15 11:16:20 Hi Marcin, Liam, koen Mar 15 11:16:26 hey do13_ Mar 15 11:19:54 lrg: touch works (same probs as the previous driver) Mar 15 11:20:21 hey lrg Mar 15 11:20:28 hey do13_ Mar 15 11:20:36 hey XorA Mar 15 11:22:07 lrg: any ideas on the pendown issues? Mar 15 11:24:12 hi, folks. I have already build e2fsprogs packages. How can I install e2fsprogs-libs into staging directory ? Mar 15 11:26:41 hey XorA Mar 15 11:27:03 do13: Did you try all the possible mask settings 0 -3 ? Mar 15 11:29:11 zecke: http://dominion.kabel.utwente.nl/koen/OE/people/people.html Mar 15 11:29:58 lrg: not yet. I haven't played with the mask settings. First I'll look at the spurious pendown events Mar 15 11:30:26 koen: nice Mar 15 11:31:28 * koen hugs awk Mar 15 11:32:15 do13: I'll try and send a patch tonight with some extra debug. imho, it may be possible for some lcd noise to cause the spurious pen downs. I'm going to ask a hw about this. Mar 15 11:32:29 hw enginner Mar 15 11:32:37 engineer Mar 15 11:32:43 koen: I would do a table from developer list: KEY | Name | IRC nick | contact Mar 15 11:33:47 hrw|work:
    is shorter as and I'm lazy Mar 15 11:33:51 patches welcome :) Mar 15 11:33:55 ;) Mar 15 11:34:27 lrg: thx. But I suspect this has something todo with the adc reading. If I disable the adc reading, there are no spurious pendown events. Mar 15 11:34:48 hi, koen . How can I install a *.ipk into staging directory ? Mar 15 11:35:38 leoncamel: you can unpack it with ar -x, but it should have installed everything needed by itself Mar 15 11:35:56 do13: ok, I'd forgotten that Mar 15 11:36:17 koen, hmm, is there anything like "prefix" I can use ? Mar 15 11:36:59 if you're on debian you can just do 'dpkg -X file.ipk ' Mar 15 11:37:15 koen, OK .. got it . Mar 15 11:37:18 lrg: playing with the mask settings makes no sense with this bug Mar 15 11:39:18 ibot: botmail for reenoo: Did you mean something like this: http://dominion.kabel.utwente.nl/koen/OE/people/people.html ? Mar 15 11:39:49 do13: i agree, I had forgotten about the aux adc. mask will only help with noise. Mar 15 11:41:04 lrg: btw. sound via headphone works (one channel, I need to control the external switch). Mar 15 11:42:50 do13: cool, we now need to get both channels working. I'll be able to debug this next week. The left & right pcm playback switches should both be unmuted as well as the headphone unmute. Mar 15 11:43:17 koen: yours html skills scares me Mar 15 11:43:34 heh Mar 15 11:44:06 I stole the framework from the OE site :) Mar 15 11:44:17 koen, I seems the "-msoft-abi=softfp" can't work here for iwmmxt . Mar 15 11:44:33 koen:
  • content
    - argh...
  • content
  • is correct Mar 15 11:45:04 lrg: yep. But there is an external switch. Have a look at the schematics. Mar 15 11:47:41 do13: I've not implemented any external switches yet, although if you know whats needed I'll happily take a patch :) Mar 15 11:49:37 ~seen daka Mar 15 11:49:42 daka is currently on #oe #nslu2-linux #openjtag #openslug, last said: 'you can use ipk-files from one to the other, if they use the same glibc etc. but files will be in the wrong place, because of the "file system setup"'. Mar 15 11:50:56 * france is back (gone 00:53:27) Mar 15 11:51:34 koen: http://www.hrw.one.pl/tmp/people.html Mar 15 11:54:52 <_law_> hrw|work, realname Patrick Steiner Mar 15 11:56:57 ok Mar 15 12:04:32 hrw|work: that's looking much better Mar 15 12:12:15 Mar 15 12:12:23 Frederic Devernay is frederic@? Mar 15 12:12:50 hrw|work: did you want something? Mar 15 12:13:08 DaKa: realname Mar 15 12:13:40 David Karlström Mar 15 12:13:43 thx Mar 15 12:14:00 but, already there... Mar 15 12:17:11 what about a founders sections? Mar 15 12:17:43 zecke: patches welcome? Mar 15 12:18:37 zecke: right, gotta give props to kergoth Mar 15 12:19:21 and schurig Mar 15 12:19:36 oh.. I was thinking that koen added them to core team and contentrated on devel list Mar 15 12:20:04 zecke: effem = Frans Meulenbroeks, JustinP = Justin Patrin, lamikr = Mika Laitio, noodles = Jonathan McDowel Mar 15 12:20:18 pb__: I just took the output of 'monotone list keys' Mar 15 12:20:39 as the note on top indicates: 'This page lists all people with write access to the OE metadata repository' Mar 15 12:21:31 phillipe = Phillipe DeSwert Mar 15 12:21:51 yuelcelm = Mustafa Yuecel Mar 15 12:22:08 cbrake = Cliff Brake Mar 15 12:22:11 Mar 15 12:22:15 Matthew Reimer = mreimer Mar 15 12:22:19 and iird ade = Adrian Davey Mar 15 12:23:03 koen: updated Mar 15 12:23:36 few people missing Mar 15 12:26:51 tmbinc = Felix Domke Mar 15 12:27:32 peteru = Peter Urbanec Mar 15 12:31:11 added founders Mar 15 12:32:00 rock Mar 15 12:32:03 hrw|work: in that case you need to adjust the description at the top Mar 15 12:33:16 first some look changes Mar 15 12:33:47 br1 = Bruno Randolf (iirc) Mar 15 12:37:14 correct Mar 15 12:37:41 could the titles be left justified? Mar 15 12:37:54 could Mar 15 12:39:01 dyoung jp30 ka6sox mwester mgoebl are unidentified Mar 15 12:39:19 * france is away: Away Mar 15 12:39:42 I have the real names for mwester and mgoebl in my mailbox Mar 15 12:39:49 I suppose Mar 15 12:39:58 but as usual thunderbird's search is useless Mar 15 12:43:05 hrw|work: it seems you removed reenoo Mar 15 12:43:10 ops.. Mar 15 12:43:18 will revert him Mar 15 12:43:21 heh Mar 15 12:44:12 mwester and mgoebl only left Mar 15 12:44:51 both are nslu2 people Mar 15 12:46:27 mgoebl left and dyoung-away does not remember him too now Mar 15 12:47:26 hrw|work: mgoebl -> Matthias Goebl Mar 15 12:47:31 gracias Mar 15 12:49:50 hrw|work: google says: mwester -> Mike Westerhof Mar 15 12:51:02 do13_: all people has info now Mar 15 12:53:31 hrw|work: there aren't any keys for schurig and kergoth in the OE databases Mar 15 12:53:47 koen: did not know Mar 15 12:54:08 and as a result they both have no write access Mar 15 12:56:24 ok. static version done Mar 15 12:58:23 I'm not sure if my browser sucks or if nails name is incorrect. Should be Øyvind Repvik Mar 15 12:58:55 I get R^vind Mar 15 12:59:09 instead os yvind Mar 15 13:00:10 http://oe.handhelds.org/people.php - feel free to improve Mar 15 13:00:19 I will add it into left menu Mar 15 13:04:49 http://oe.handhelds.org/contact.php has contributors list which probably should be included in people.php Mar 15 13:06:45 the contributors could have their own section in people.php Mar 15 13:08:41 hi everyone Mar 15 13:09:49 hrw|work: don't forget to mail oe@ Mar 15 13:10:31 yep Mar 15 13:10:48 aargh Mar 15 13:10:53 ~lart pdflatex Mar 15 13:10:53 * ibot decapitates pdflatex conan the destroyer style Mar 15 13:39:22 pH5: I made some changes to angstrom to enable eabi for arm and to set the correct providers, could you see if that works for you? Mar 15 13:40:40 koen: except for the TARGET_OS_arm thingy I had this configuration already. Mar 15 13:40:50 pH5: for the kernel headers: http://pastebin.com/603401 Mar 15 13:42:08 koen: there's a rc6 Mar 15 13:42:49 koen: Did you succeed to build glibc against the vanilla kernel headers? Mar 15 13:43:13 only for the old abi Mar 15 13:51:44 My glibc finally finished packaging. Before recompiling the whole lot again, I'll continue and try if I can build a boostrap-image. Mar 15 13:56:09 NOTE: package zbedic-0.9.9+1.0rc1-r1: task do_build: completed Mar 15 13:57:08 pH5: don't be afraid to push your changes, since glibc+gcc are broken anyway Mar 15 14:12:00 ~seen mickeyl Mar 15 14:12:08 mickeyl was last seen on IRC in channel #oe, 19h 59m 1s ago, saying: 'bbiab'. Mar 15 14:12:56 new definition of 'a bit' ;D Mar 15 14:13:20 he said he would be on irc this afternoon Mar 15 14:13:55 I know Mar 15 14:14:01 I want to takeover zbedic Mar 15 14:15:54 re Mar 15 14:16:03 zecke: wb Mar 15 14:16:06 hey zecke Mar 15 14:25:04 koen: help Mar 15 14:25:09 koen: you built busybox against glibc 2.4 yesterday? ping6.c is missing the definition of ICMP6_MEMBERSHIP_REDUCTION in include/netinet/icmp6.h Mar 15 14:25:30 pH5: I said that I couldn't build busybox Mar 15 14:25:38 I got the same error Mar 15 14:25:47 maybe busybox 1.1.0 works? Mar 15 14:25:52 hrw bugzillad the recipe Mar 15 14:26:40 koen: oh, ok. I'll have a look. Mar 15 14:28:30 hrw|work: hi there Mar 15 14:28:34 did I lose something? 15:21 - 15:30 Mar 15 14:28:37 brb Mar 15 14:29:02 hrw|work: no, you didn't loose anything. Mar 15 14:31:42 hrw|work: Looking at the list of people btw.. The capitalization of my nick is incorrect, and the first character of my name should probably be replaced with "Ø" instead of the Ã(whatever) characters ;) Mar 15 14:32:02 ok Mar 15 14:32:44 koen: http://sources.redhat.com/cgi-bin/cvsweb.cgi/libc/inet/netinet/icmp6.h.diff?r1=1.9&r2=1.10&cvsroot=glibc&f=h Mar 15 14:32:58 koen: it seems some defines were renamed Mar 15 14:33:18 Oyvind Repvik slashed Mar 15 14:33:19 indeed it does Mar 15 14:33:48 NAiL: copied from non-utf8 terminal.. Mar 15 14:34:19 thank you :) Mar 15 14:35:54 hrw|work: zecke's keyid is freyther@ Mar 15 14:37:47 how to describe pb_? Mar 15 14:38:42 All-round masterhacker? Mar 15 14:40:56 <_law_> hmm whats my function in OE ? Mar 15 14:41:07 <_law_> compiler ;-) Mar 15 14:45:43 koen: NOTE: package busybox-1.01: completed Mar 15 14:45:43 http://en.pastebin.ca/45806 Mar 15 14:45:56 pH5: 1.01 != 1.1.0 Mar 15 14:46:51 pH5: that patch looks ok Mar 15 14:47:46 hrw|work: with 1.1.0 it is exactly the same. Mar 15 14:47:53 hrw|work: Is a suitable function for me, Laptop Replacement Developer? thats my interest in OE :-) Mar 15 14:48:54 XorA: ;D Mar 15 14:49:04 hrw|work: or mplayer video watcher :-) Mar 15 14:49:28 during evening I maybe prepare better looking version Mar 15 14:49:46 Pr0no Digital Assistant Mar 15 14:50:34 XorA: P0rn Doubler Analyzer Mar 15 14:50:49 or P0rn Doubler Assistant Mar 15 14:51:05 re zecke Mar 15 14:51:31 thanks again Mar 15 14:52:53 koen: will building the eabi kernel with gcc-cross-3.4.4 work? Mar 15 14:53:23 pH5: I don't think so Mar 15 14:53:39 pH5: is 3.4 still needed for hh.org 2.6? Mar 15 14:54:28 pH5: also note that kernels need a seperate defconfig for EABI Mar 15 14:54:46 pH5: since the compatibility patches aren't in yet Mar 15 15:03:19 ~lart linear variance estimation and null-space error minimization Mar 15 15:03:19 * ibot shoots linear variance estimation and null-space error minimization in the head Mar 15 15:06:02 * koen decides it's time to have a beer Mar 15 15:11:50 damn, 3105-1 doesn't apply cleanly against 2.6.15-hh0 Mar 15 15:13:13 http://oe.handhelds.org/people2.php looks better? Mar 15 15:22:17 hrw|work: I think it does Mar 15 15:25:51 replaced Mar 15 15:28:37 got first change request Mar 15 15:31:27 now we just need to teach bitbake to generate a html maintainer list Mar 15 15:34:54 03florian 07org.oe.dev * r15add738... 10/packages/gpe-keylock/gpe-keylock_0.12.bb: gpe-keylock: add 0.12 Mar 15 15:34:59 03florian 07org.oe.dev * r9cd0f34b... 10/conf/bitbake.conf: bitbake.conf: Update default GPE mirror. Mar 15 15:35:04 03florian 07org.oe.dev * r914aee76... 10/conf/bitbake.conf: disapproval of revision '9cd0f34b9dc26c55cd6929716454080ef9f93526' Mar 15 15:35:08 03florian 07org.oe.dev * r6bc05c2a... 10/packages/keylaunch/ (4 files in 2 dirs): keylaunch: add 2.0.10, drop old versions. Mar 15 15:35:16 03florian 07org.oe.dev * r4d3dc9c8... 10/packages/display-brightness/ (display-brightness.sh display-brightness_1.0.0.bb): display-brightness: Add 1.0.0 contributed by Patrick Steiner. Mar 15 15:35:22 03florian 07org.oe.dev * r4c803688... 10/packages/keylaunch/files/ (7 files in 7 dirs): keylaunch: Add keylaunchrc for multiple other clamshell Zaurus models. Please test! Mar 15 15:43:26 03florian 07org.oe.dev * r9a62f32e... 10/packages/gpe-filemanager/ (gpe-filemanager_0.20.bb gpe-filemanager_0.25.bb): gpe-filemanager: add 0.25, drop obsolete releases Mar 15 15:49:38 hmm.. keylaunch 2.0.10... Mar 15 15:51:17 hi mickeyl Mar 15 15:51:21 hi Mar 15 15:52:05 mickeyl: I want to takeover zbedic from you - any objections? Mar 15 15:52:13 hrw|work: feel free Mar 15 15:52:38 1.0rc1 was released Mar 15 15:53:51 how goes world build? Mar 15 15:55:32 progressing.... i have no idea if that goes alphabetically, but it's doing gettext atm. :D Mar 15 15:55:43 hi mickeyl Mar 15 15:56:23 03florian 07org.oe.dev * rab0a2d21... 10/packages/keylaunch/keylaunch_2.0.10.bb: keylaunch: add missing 2.0.10 bb Mar 15 15:56:24 hi do13 Mar 15 15:56:30 how's tosa progressing ? Mar 15 15:57:03 hrw|work: a depressing number of packages don't build on a world build Mar 15 15:57:07 hrw|work: but that's another story... Mar 15 15:57:36 hi mickey Mar 15 15:57:37 mickeyl: can you log them? Mar 15 15:57:43 hi pb_ Mar 15 15:57:50 mickeyl: got new touch and sound driver Mar 15 15:58:10 hrw|work: sure, i will run another world build after this one finishes and pipe the complete output into a file. Mar 15 15:58:17 do13: excellent Mar 15 15:58:20 hi drw, how are things? Mar 15 15:58:21 mickeyl: fine Mar 15 15:59:14 hi mickeyl Mar 15 15:59:27 and hrw|work, etc. Mar 15 15:59:40 mickeyl: doing good Mar 15 15:59:51 mickeyl: I changed qte to build threaded lib. can you look at the patch http://www.do13.de/oz/temp/qte_mt.patch Mar 15 16:00:24 hi drw Mar 15 16:01:51 mickeyl: this creates an additional lib which can be used for konqe Mar 15 16:02:46 do13: looks good so far, feel free to commit to .dev. One of these days we can try to change the qmake default config to +=thread to see what blows up when opie apps link against the multithreaded lib Mar 15 16:03:35 mickeyl: But these lib is 900K more then the old one Mar 15 16:03:51 hmm do we want to kill qte_2.3.10 Mar 15 16:04:03 do13_: that's unfortunate, but konq is not lightweight either. Mar 15 16:04:23 do13_: I would not do that for Konq now Mar 15 16:04:34 zecke: ? Mar 15 16:04:35 do13_: I have a threading lib that could be shared by OpiePlayer and Konq Mar 15 16:04:48 zecke: he's not killing the unithreaded, he's just adding a threaded lib. Mar 15 16:04:59 +multi Mar 15 16:05:13 zecke: libtq + libqte-mt Mar 15 16:05:47 zecke: oh you mean an additional lib on top of the non-mt qte? Mar 15 16:10:34 bbl Mar 15 16:10:36 mickeyl: the diff looks like -qte_2.3.10 Mar 15 16:10:39 cya Mar 15 16:11:05 zecke: -qte_2.3.10 +qte.inc Mar 15 16:11:33 sorry *too much input at once* currently Mar 15 16:11:45 -qte2 + qt4 Mar 15 16:11:45 happens Mar 15 16:12:03 koen: thats for future Mar 15 16:12:14 oh, hooray. gpe-image built successfully w/ gcc-4.0.2 cross Mar 15 16:12:22 the gcc-4.1 wasn't so happy though Mar 15 16:12:24 pthread errors Mar 15 16:13:09 pH5: could you push that busybox patch? Mar 15 16:21:26 emte: ftp working via http:// Mar 15 16:22:24 eep Mar 15 16:22:48 wpa_supplicant-0.5.1 still wants to build against openssl.... fixing Mar 15 16:23:16 people, please report breakage of your favorite ssl-using app so we can get it gnutls-ready Mar 15 16:23:52 so far we've discussed wpa_supplicant, sylpheed, and others Mar 15 16:24:20 jnc: sylpheed 2.2.X should no longer need gnutls patches Mar 15 16:24:36 hi all :) Mar 15 16:24:37 XorA: that's fantastic news, thank you for the original gnutls patch btw Mar 15 16:24:47 i think it was attributed to you, from the posts i saw Mar 15 16:24:57 jnc: I disown that patch, it was horrible Mar 15 16:25:01 * jnc giggles Mar 15 16:25:04 yeah, it was Mar 15 16:25:04 jnc: ntp? Mar 15 16:25:10 a fast question, what the nema of 'new' graphic interface amstrong ??? where i can find screenshoot ? Mar 15 16:25:10 ntp... okay adding to the list Mar 15 16:25:30 jnc: cherokee should work out of the box with gnutls Mar 15 16:26:05 jnc: but I had nothing to do with the openssl compatability in gnutls, so dont get the two mixed up Mar 15 16:26:14 ohh Mar 15 16:26:21 yeah i had those mixed up. noted. Mar 15 16:26:36 gremlin[it]work: angstrom is a distribution, not an environment Mar 15 16:26:47 jnc: and I mean new versions of sylpheed should no longer need the openssl patches to gnutls, it should support gnutls directly Mar 15 16:27:12 gremlin[it]work: you probably mean http://vanille.de/temp/angstrom-bootmanager.png and http://vanille.de/temp/angstrom-calibrate.png Mar 15 16:27:24 gremlin[it]work: which is just the installer for the distribution Mar 15 16:28:06 ah ok ... but it's based on e17 or aren't related ? Mar 15 16:28:40 the installer and the bootmanager are likely to be based on EFL Mar 15 16:28:47 XorA: any reason you know of we should be keeping older versions of sylpheed prior to 2.2.x around? Mar 15 16:28:59 jnc: only that Im not the maintainer Mar 15 16:29:19 jnc: ntp is an example of ssl app but we rather do not use ssl powered ones Mar 15 16:29:48 jnc: we have konqueror-embedded linked to openssl (but it is crap util) Mar 15 16:30:14 i'm a little nervous about changing apps that aren't in my focus. i'll be bugging maintainers Mar 15 16:30:19 a lot Mar 15 16:30:28 03jnc 07org.oe.dev * r4d91686a... 10/packages/wpa-supplicant/files/defconfig-gnutls: change defconfig of 0.5.1 to build against gnutls instead of openssl. Mar 15 16:30:54 jnc: change it, send patch to bugtracker, pester maintainer Mar 15 16:31:36 are there example platforms for which no SSL is wanted or used? Mar 15 16:31:56 basically we rather avoid openssl as much we can Mar 15 16:32:02 OE seems to be a case of choosing "sane defaults" unlike Gentoo where you'd just throw a use flag in for something Mar 15 16:32:16 oh not openssl, i mean any SSL. since gnutls may be a workable alternative Mar 15 16:32:17 pb__: and _law_ are the listed sylpheed maintainers Mar 15 16:32:24 jnc: OE lack USE flag and we know it Mar 15 16:32:33 yeah Mar 15 16:32:42 i think it's simpler without USE flags, i like that Mar 15 16:32:44 USE_FLAGS suck Mar 15 16:32:52 it breeds incompatibility Mar 15 16:32:53 hrw|work: not that USE flags would scale very good Mar 15 16:33:34 USE_FLAGS are OK if you have no concept of packages or distributing, like gentoo Mar 15 16:33:35 hi, koen , what is angstrom ? Mar 15 16:33:44 leoncamel: a new distribution Mar 15 16:34:07 will have to finish wpa_supplicant fixes later tonight. work time Mar 15 16:34:51 koen, something like familiar ? Mar 15 16:35:10 leoncamel: much better and actually maintained Mar 15 16:35:40 koen, where can I find it ? Mar 15 16:35:49 in oe :) Mar 15 16:35:59 hehe Mar 15 16:36:57 koen, no web site, and no help ? Mar 15 16:37:15 leoncamel: it is Work In Progress?! Mar 15 16:37:38 leoncamel: http://ewi546.ewi.utwente.nl/angstrom/ is all we got Mar 15 16:37:50 leoncamel: I think currently it needs more developers than people testing it Mar 15 16:38:12 hmm, its bootloader is more fancy .. :) Mar 15 16:38:22 leoncamel: no Mar 15 16:38:27 leoncamel: bootmanager != bootloader Mar 15 16:38:40 in case someone says FUD: familiar is unmaintained until http://handhelds.org/hypermail/familiar-dev/current/0753.html gets answered Mar 15 16:39:07 leoncamel: we have been planning this thing since autumn last year and since the people working on it are the same people that were responsible for the work on OZ, OpenSIMpad, and Familiar 0.8.x, we have been finishing the lastest sets of distributions in the last months before we actually could start with some more defined processes. Mar 15 16:40:02 zecke, what is the difference ? Mar 15 16:41:11 mickeyl, hmm, why not familiar? and why start this new distribution ? Mar 15 16:41:35 leoncamel: have a look at the last couple of OZ newsletters, the reasons are layed out there. Mar 15 16:41:45 leoncamel: well, as this is a developer channel of Embedded devices, I would assume that you know what a Bootloader is? Mar 15 16:41:47 a new name was necessary because of political reasons Mar 15 16:41:54 leoncamel: think of it as BIOS for your pda? Mar 15 16:41:55 s/necessary/wanted/ Mar 15 16:41:57 whatever :D Mar 15 16:42:03 leoncamel: familiar is unmaintained at the moment and a dead end Mar 15 16:42:48 hmm i prefer to think about it as taking a different direction :) Mar 15 16:42:59 if that's a dead end, we don't know Mar 15 16:43:06 * mickeyl diplomatic today Mar 15 16:43:10 right Mar 15 16:43:18 maaah Mar 15 16:43:22 koen, well, the angstrom is next generation of familiar ? Mar 15 16:43:40 There's like a gazillion mails on the OE list every day :\ Mar 15 16:43:56 NAiL: better that way than dead :) Mar 15 16:44:02 NAiL: where gazillion is 10^verybig? Mar 15 16:44:12 NAiL: silent mailing lists are depressing me Mar 15 16:44:20 gazillion ~= toomuch ;) Mar 15 16:44:36 heh Mar 15 16:44:37 mickeyl: well, yeah. But I should be able to keep up too ;) Mar 15 16:45:07 * NAiL should remove some mailinglists. And filter some others... Mar 15 16:45:18 NAiL: you could also add the gmane newsfeed to your news reader Mar 15 16:45:32 NAiL: makes it easier to ignore :) Mar 15 16:45:36 hehe Mar 15 16:45:55 * koen has LAK as nntp feed Mar 15 16:46:30 LAK doesn't bother me much, since most of the traffic there is obviously not connected to anything I'm doing ;) Mar 15 16:46:35 mickeyl: millions of bug mails a day are... Mar 15 16:47:19 zecke: true. does bugzilla feauture some kind of digest? It would make sense to group bug message referring to the same bug. Mar 15 16:47:28 s/message/messages/ Mar 15 16:48:00 mickeyl: also OZ bugs should be sent to oz-devel, not oe@ Mar 15 16:48:41 I guess it's time for me to check if there's any bugs I can fix too. Mar 15 16:49:26 koen: historically, we suck at seperating all those projects we work on. I hope that we can improve that in the future. Mar 15 16:50:08 mickeyl: historically I get yelled at because OZ=OE=bugmails to oe@ Mar 15 16:51:18 hmm Mar 15 16:51:31 why are b ugs like oe bug #26 still open? Mar 15 16:51:41 http://bugs.treke.net/show_bug.cgi?id=26 Mar 15 16:52:34 laziness Mar 15 16:52:48 03pH5 07org.oe.dev * rfd7121b0... 10/packages/busybox/ (files/glibc2.4-icmp6.patch busybox_1.01.bb): busybox-1.01: make ping6.c compile against glibc-2.4 Mar 15 16:52:49 because we have too few people being interested in uclibc Mar 15 16:53:06 re Mar 15 16:53:08 because we have not enough automated regression tests yet Mar 15 16:53:10 etc. Mar 15 16:53:12 do13: wb Mar 15 16:53:30 pH5: did you manage to make a patch for the mpfr problem? Mar 15 16:54:54 zecke: what is your objection for libqte-mt (2.3.10)? Mar 15 16:54:56 koen: no, I did it your way :) Mar 15 16:55:35 do13: s/objection/fear/ Mar 15 16:55:54 do13: I fear that one can not easily select between a multi threaded and none multi threaded version of QtE Mar 15 16:56:46 i guess we will build both and apps will choose by defining CONFIG appropriately Mar 15 16:56:56 (at first) Mar 15 16:57:18 not both :} Mar 15 16:57:32 why not? that's a seamless transition. Mar 15 16:57:43 that is a waste of flash Mar 15 16:57:48 ~4 mb Mar 15 16:57:49 that way people can install the new konqueror to existing Opie Mar 15 16:58:12 i wouldn't call it a waste. people have large SD cards nowadays. Mar 15 16:58:16 memory is another issue though Mar 15 16:58:30 and we can provide konqe for 3.5.4 as an upgrade Mar 15 16:58:38 what works is a ln -s libqte-mt.so libqte.so Mar 15 16:58:43 (not the other way around) Mar 15 16:58:44 like i said, we can try later to see which parts of Opie blowup when we change everything to link against the multithreaded lib Mar 15 16:58:53 none Mar 15 16:59:06 optimits Mar 15 16:59:09 optimist, even Mar 15 16:59:26 zecke: did you tried this yet? Mar 15 16:59:34 do13: which part? Mar 15 16:59:37 i know you probably tried that, but since the blowup from 2.3.9 to 2.3.10, i'm afraid of _every_ QtE change. Mar 15 16:59:54 opie with libqte-mt Mar 15 17:00:02 mickeyl: a) Wimpie did that b) ROAD does it for KonqE Mar 15 17:00:19 mickeyl: c) QtE MT just add a qApp->lock() and qApp->unlock() method Mar 15 17:00:34 mickeyl: d) I was there when KDE switched from Qt2 to Qt3 mt Mar 15 17:01:10 but we need testing nevertheless Mar 15 17:01:13 zecke: whats about rtti and exeptions? Mar 15 17:01:27 do13: it needs rtti for sure Mar 15 17:01:52 do13: and exceptions as well. Actualy exceptions should just be on on DOM/KHTML Mar 15 17:03:25 3.5.4 is out anyway, so the roadmap is clear. we want to provide an upgrade for existing users first, and then we can think about switching to qte/MT as the one and only for the next release. Mar 15 17:03:36 (imo) Mar 15 17:03:50 s/users/installations/ Mar 15 17:03:58 * mickeyl goes back to focus on the thesis Mar 15 17:03:59 cu Mar 15 17:08:50 cu Mar 15 17:40:15 yeah i saw your changes in #gpe, thanks france Mar 15 17:44:02 any idea on why bitbake is complaning about /etc/passwd ? Mar 15 17:45:41 elaborate? Mar 15 17:46:56 it says it cannot find rootfs/etc/passwd Mar 15 17:47:13 Thats just a tad different then /etc/passwd Mar 15 17:48:10 hmm sorry, this is my first attempt to compile oe Mar 15 17:50:43 I don't know anything about how GPE works, but it probably sounds like your're trying to run zap_root_password but not installing base-passwd Mar 15 17:51:45 yep Mar 15 17:52:05 zap_root_password is where is it trying to append to file Mar 15 17:52:20 but in my work dir i can see base-passwd Mar 15 17:52:29 bluebugs, there are a couple people who have suddenly come up against this bug again Mar 15 17:56:09 yay, "NOTE: package handhelds-pxa-2.6-2.6.15-hh0+cvs20060315: completed" Mar 15 17:56:16 bluesign: rebuild base-passwd Mar 15 17:56:22 pH5: with eabi patches? Mar 15 17:56:27 koen: how can I do that? Mar 15 17:56:31 koen: yess Mar 15 17:57:11 bluesign: bitbake -i rebuild base-passwd Mar 15 17:57:14 pH5: cool Mar 15 17:57:26 pH5: is TARGET_OS in overrides? Mar 15 17:57:39 pH5: or should we just use _angstrom overrides? Mar 15 17:58:25 koen: and three fixups for gcc 4.1, I think. it doesn't like static function declarations inside other functions, as it seems. Mar 15 17:58:42 koen: what do you mean? both angstrom and TARGET_OS are in overrides here. Mar 15 17:58:49 ah, cool Mar 15 17:59:24 that means we can add EABI patches with SRC_URI_append_none-linux-eabi = "scarypatch;patch=1" Mar 15 18:02:07 koen: like this: http://en.pastebin.ca/45828 ? Mar 15 18:03:55 pH5: like that :) Mar 15 18:04:28 pH5: although you can leave out old-abi-compat* in this override Mar 15 18:05:29 koen: NOTE: Not creating empty archive for base-passwd-dev-3.5.9-r2 Mar 15 18:05:29 (is it something bad?) Mar 15 18:05:44 bluesign: that one is harmless Mar 15 18:26:50 pH5: do you have a tarball or .bb for the hacked kernel headers? Mar 15 18:27:04 re Mar 15 18:27:12 zecke: wb Mar 15 18:36:19 koen: I'll upload the tarball, give me some minutes. Mar 15 18:41:32 I need to define ac_cv_linux_vers in site/ for libpcap. should I just add this one value to a new site/arm-none-linux-gnueabi, or should I start from a copy of an existing file? Mar 15 18:42:55 pH5: I suspect you can copy the arm-linux one, but to be sure ask pb__ or pbrook or drow Mar 15 18:44:29 zecke: how useable is the simpad 2.6 kernel? Mar 15 18:44:38 pb__: Is it safe to cp site/arm-linux site/arm-none-linux-gnueabi ? Mar 15 18:46:10 koen: in bad shape I guess, that depends on how well the hh.org merges went Mar 15 18:47:34 zecke: so you can make guylhem do the tedious work of getting it to work again? Mar 15 18:47:54 koen: where do you get that from? Mar 15 18:48:07 koen: I think guyhlem should stick with 1.3 Mar 15 18:48:12 koen: or 2.0 Mar 15 18:49:17 NOTE: package bootstrap-image-1.0: completed Mar 15 18:49:22 can't help it, I have to flash that Mar 15 18:49:38 hehe Mar 15 18:50:20 pH5: do you where your "Prüfungs-Ausschuss" (Physic Department) is? Mar 15 18:50:36 pH5: I'm starting Physics as "Nebenfach" Mar 15 18:52:32 Good evening all Mar 15 18:53:05 hi Mar 15 18:53:29 I have a special question ... Mar 15 18:54:40 we (approx 10 developers) want to set up a bitbake environment with a central cross / staging area and also some kind of a local staging area. deploy should be global or local. Mar 15 18:54:48 Is that somehow possible ? Mar 15 18:55:18 zecke: cool. the pruefungsausschuss consists of profs. brewer, bosse and schwendtner I think. I don't know the name of the new secretary, but I think the place you want to go to is room 1.1.14a. Mar 15 18:55:19 hi uv1 Mar 15 18:56:01 pH5: I have a "Nachteilsausgleich" to take oral exams instead of written ones for Computer Science, I need to 'migrate' that Mar 15 18:57:08 uv1: well, having a shared cross/staging area is easy. making it know to not rebuild all those things if another developer has already done so is something else entirely ;) Mar 15 18:57:41 kergoth_: Any ideas how to do so ? Mar 15 18:58:20 uv1: do you want that everyone can rebuild cross/? Mar 15 18:59:06 uv1: e.g. you could have a dedicated 'cross build environment' and set it as ASSUME_PROVIDED Mar 15 18:59:09 Not necessarily. I just have developers, which use different staged stuff than other, which e.g. build distributions on stable packages. Mar 15 18:59:35 zecke: Sure cross is not that problem. The staging area is in fact. Mar 15 18:59:36 uv1: is it a matter of speed? Mar 15 19:00:19 koen: http://userpage.fu-berlin.de/~zabel/hx4700/eabi/linux-libc-headers-2.6.15.99.tar.bz2 Mar 15 19:00:47 a good portion of staging is target architecture stuff, which is populated by the same packages that emit your ipks or whatever. so it wouldnt make sense to make that part of staging global unless you make teh deployed packages for those global as well Mar 15 19:00:54 zecke: Yes and No, Speed, Space and lets say trying new things while building distro is stable Mar 15 19:01:35 uv1: speed: use icecream :) Mar 15 19:01:46 uv1: 10 people would build a pretty nice build cluster Mar 15 19:02:01 kergoth_: Deployed packages can and should be global at the end to go into the distro anyway. Mar 15 19:02:38 uv1: you can set the CROSS DIR and DEPLOY dir to a global directory Mar 15 19:02:46 dont forget to move stamps. Mar 15 19:02:51 those wouldnt make since without those Mar 15 19:02:57 sense, even Mar 15 19:05:07 kergoth_: ? Mar 15 19:05:40 uv1: global stamp dir Mar 15 19:07:36 Sure, the idea is to have a "global" bitbake run (e.g. bootstrap-image) and on top of that build several development extensions until they are deployable. Modify the global image bb files, rebuild global, remove all locals and start from the beginning. Mar 15 19:10:06 Imagine you have a gcc-4.1 system in test and your stable system uses gcc-3.4.4 (glibc ...). It just would be nice not to do everithing 10 times ;-( Mar 15 19:11:58 BTW : Is the cross dir "installable" to somewhere else ? Mar 15 19:12:07 uv1: the hard part is managing the locals as well as the globals, when bitbake is only aware of a single stamps and staging. it wouldtn know how to build a package and store its stamps locally, but check that against the global stamps for dependencies Mar 15 19:12:18 with a recent toolchain, it should be relocatable yes Mar 15 19:12:41 kergoth_: recent means 3.4 + ? Mar 15 19:13:16 dont recall exactly when it was fixed Mar 15 19:13:34 3.4 will definately be relocatable, just not sure what older ones are Mar 15 19:14:00 kergoth_: 3.4 is old enough. Mar 15 19:14:12 hehe Mar 15 19:15:09 Thanks all for audience and for help. I will try to setup an interims solution and then dive into the bb system to possibly find a solution. Mar 15 19:18:10 pH5: I think it should be, pretty much. Might want to eyeball it to look for things that are different. Mar 15 19:18:42 Bye ... Mar 15 19:19:01 hey pb_ Mar 15 19:19:26 yo kergoth Mar 15 19:26:10 koen: It actually boots, and I even got the 5-minute-hang during the mounting of local file systems. Mar 15 19:26:19 there are some errors, though: http://en.pastebin.ca/45835 Mar 15 19:30:21 evening Mar 15 19:32:52 pH5: yay! Mar 15 19:37:27 hey greentux Mar 15 19:37:44 hi koen... nokia/oe news? Mar 15 19:38:17 pH5: If you can push the fix for the new linuxheaders, we can clean up the small bits tomorrow/weekend Mar 15 19:38:27 greentux: no real nokia new Mar 15 19:38:42 nokia? what are they up to? Mar 15 19:39:24 kergoth_: make their maemo stuff a bit more sane Mar 15 19:39:32 ah. Mar 15 19:40:00 which means we can update the maemo bits in OE :) Mar 15 19:40:29 cool :) Mar 15 19:40:37 * kergoth_ tries to think of more items for the old http://oe.handhelds.org/cgi-bin/moin.cgi/Buildsystems_2fRulesOfThumb doc Mar 15 19:42:04 hmm, items 6 and 7 should be made more generic in that doc Mar 15 19:45:10 koen: I can't. This is what my glibc_2.4.bb looks like: http://en.pastebin.ca/45837 Mar 15 19:51:44 I want to compile fbvnc (a frame buffer VNC client that is part of the PocketWorkstation). Can I use OE to compile it? How? (Or how can I learn how?) Mar 15 19:55:55 re Mar 15 19:57:55 03yuecelm 07org.oe.dev * r9745eec9... 10/packages/ (davfs2/davfs2_0.2.7.bb subversion/subversion_1.2.0.bb): Mar 15 19:58:00 davfs2-0.2.7,subversion-1.2.0: correct version info in DEPENDS Mar 15 19:58:02 - brackets only works in RDEPENDS Mar 15 19:58:04 - remove --with-ssl configure option (inherit abilities from neon) Mar 15 20:00:53 koen: it says " Nothing provides dependency libgpg-error0" Mar 15 20:00:57 koen: any idea? Mar 15 20:01:42 bluesign: no idea on that Mar 15 20:02:06 koen: k thanks Mar 15 20:02:34 bluesign: s/libgpg-error0/libgpg-error in DEPENDS probably Mar 15 20:03:43 ~lart wlags modules Mar 15 20:03:43 * ibot takes wlags modules to the vet for a "special" visit Mar 15 20:03:48 ~snack Mar 15 20:03:51 * ibot snack? yes yeaaas yeaaaaaaaaaaas pleaaase :P Mar 15 20:09:02 pH5: I've pushed a slight variation of your glibc .bb Mar 15 20:10:07 ~lart patch for not finding existing file Mar 15 20:10:07 * ibot steals patch's mojo for not finding existing file Mar 15 20:10:09 03koen 07org.oe.dev * r5358ebbc... 10/packages/glibc/glibc_2.4.bb: Mar 15 20:10:10 glibc 2.4: use eabi capable headers Mar 15 20:10:12 * remove maintainer field Mar 15 20:11:50 pH5: I mirrored the headers on ewi Mar 15 20:15:51 ~lart me Mar 15 20:15:51 * ibot readies the nuke launcher and fires some rounds at koen Mar 15 20:16:42 03koen 07org.oe.dev * r8ff55ce5... 10/packages/glibc/glibc_2.4.bb: glibc 2.4: ooops, forgot to change a version number Mar 15 20:16:56 koen: heh :) Mar 15 20:21:07 03pH5 07org.oe.dev * rd8d3aaf5... 10/packages/cairo/cairo_1.0.2.bb: cairo-1.0.2: fix build for gcc4, patch from cvs Mar 15 20:21:29 * koen is thinking about gcc-cross-sdk_4.1.0.bb Mar 15 20:22:19 koen: "the adapter works fine under the Opie-Embedded installation" Mar 15 20:22:28 hah Mar 15 20:23:04 gosh Mar 15 20:23:09 03pH5 07org.oe.dev * r7e125248... 10/site/arm-none-linux-gnueabi: site/arm-none-linux-gnueabi: add new site file for EABI builds Mar 15 20:25:05 haha Mar 15 20:25:26 rebooot... Mar 15 20:27:12 pH5: | ../ports/sysdeps/arm/mp_clz_tab.c:23:40: error: sysdeps/generic/mp_clz_tab.c: No such file or directory Mar 15 20:28:22 koen: I swear, I did not have this problem. Mar 15 20:28:43 weird Mar 15 20:29:34 pH5: arm-none? Mar 15 20:29:44 re Mar 15 20:30:13 kergoth_: yes, stupid gnu put their foot down on the naming Mar 15 20:30:27 what do you mean? Mar 15 20:30:48 that the naming is messed up now Mar 15 20:30:50 goewi -XCvt screen -RDU Mar 15 20:31:05 i still dont see waht you mean. is none part of the TARGET_ARCh or the TARGET_OS? Mar 15 20:31:05 with the vendor ending up between arm and linux Mar 15 20:31:17 yes, i know about vendor, but vendor isnt used in our CONFIG_SITE variable Mar 15 20:31:17 target os Mar 15 20:31:23 .. Mar 15 20:31:28 we have a target vendor variable for a reason Mar 15 20:31:56 set TARGET_VENDOR to none and rename the site file back to what it shoudl be. Mar 15 20:32:01 er -none or whatever Mar 15 20:32:54 kergoth_: so that should just be arm-linux-gnueabi? Mar 15 20:33:13 hrw: were did you see that quote? Mar 15 20:33:30 the site file is TARGET_ARCH-TARGET_OS, so yes, if TARGET_OS is linux-gnueabi, it should be arm-linux-gnueabi Mar 15 20:33:36 the vendor belongs in the vendor, not in the os Mar 15 20:34:04 vendor was intentionally left out of the site file naming. test results arent vendor specific Mar 15 20:35:00 koen: ebay Mar 15 20:35:00 who here is building the release for collie? I'm having a problem and I was wondering if you are having it to... Mar 15 20:35:13 koen: read 'donation for poodle' on oesf for link Mar 15 20:35:20 Zero_Chaos: yes? Mar 15 20:35:30 hrw: sir. Have you tested it yet? Mar 15 20:35:42 Zero_Chaos: yes - gpe flavour Mar 15 20:35:49 Zero_Chaos: and opie was also tested Mar 15 20:36:45 hrw: can you please test opie-embeddedconsole (when unchecking "wrap" the scroll bar does not appear at the bottom) and konqueror-embedded errors and does not start, big message, half cut off. both under opie. Mar 15 20:37:02 pH5: I'll try a build on another machine Mar 15 20:37:13 Zero_Chaos: vertical scrollbar was never shown iirc Mar 15 20:37:18 hrw: also is anything being released as 3.5.4 that is not built from .oz354fam083 Mar 15 20:37:33 Zero_Chaos: but I will test 3.5.4/opie/collie now Mar 15 20:37:38 hrw: horizontal scroll bar doesn't appear when I turn off wrap, vertical works fine. Mar 15 20:37:42 alsa.bb references ${ALSA_PACKAGES} which was the reason of my problem, It is fixed in dev branch, will it be committed to stable also? Mar 15 20:37:47 hrw: thanks, just wondering if it's me. Mar 15 20:37:53 Zero_Chaos: ops. I meant horizontal Mar 15 20:38:21 hrw: it was showing about 2 weeks ago (my build from .oz354fam083) Mar 15 20:38:45 hrw: a horizontal scroll bar is rather needed once wrap is turned off, as it, to SEE the rest of the screen Mar 15 20:38:57 03pH5 07org.oe.dev * r86bf9f4c... 10/site/ (arm-none-linux-gnueabi arm-linux-gnueabi): correct arm-none-linux-gnueabi to arm-linux-gnueabi Mar 15 20:39:10 pH5: thanks Mar 15 20:40:30 btw - I finally HAVE serial cable for Z - need to resolder 9<>9 nullmodem to 9<>25 Mar 15 20:40:40 Zero_Chaos: flashing Mar 15 20:40:59 hrw: thanks... BTW, how much abuse do you think the collie's flash can take? Mar 15 20:41:27 Zero_Chaos: dont know Mar 15 20:41:38 hrw: you mean you have not reached that point yet? Mar 15 20:41:46 hrw: it has not died I mean. Mar 15 20:42:08 Zero_Chaos: still works Mar 15 20:42:25 hrw: fair enough, i'll bet yours get a lot more use than mine :-) Mar 15 20:44:12 Zero_Chaos: btw - by default we use opie-console not opie-embeddedkonsole Mar 15 20:44:33 memory_cs.o compiled... Mar 15 20:45:00 hrw: I am aware of that, I prefer opie-embeddedkonsole because it exit's when I ask, and it just seems nicer (person preferance) Mar 15 20:56:13 heh.. building kernel module without makefile is not nice Mar 15 20:56:50 hrw? Mar 15 20:57:24 Zero_Chaos: I'm building kernel module for cfvga card from IO-data and source lack makefiles Mar 15 20:58:56 iirc kdrive doesn't need the kernel driver to make the card work ;) Mar 15 20:59:34 koen: but qtopia tool need Mar 15 20:59:42 Packaged contents of cfvga into /home/hrw/devel/build/oz354-prefinal/tmp/deploy/ipk/cfvga-2.4.18-rmk7-pxa3-embedix_1.0-r0_poodle.ipk Mar 15 20:59:52 yeah;) Mar 15 21:06:33 koen: if I set TARGET_OS_arm to "linux-eabi" and TARGET_VENDOR_arm to "-none", all my build paths are wrong :-/ Mar 15 21:07:00 how so? Mar 15 21:07:25 kergoth_: ok, so they are correct now, but they were wrong before Mar 15 21:07:41 (arm-none-linux-eabi --> arm-linux-eabi) Mar 15 21:08:33 the vendor is part of TARGET_SYS. if any paths are using the component variables directly rather htan using TARGET_SYS, they're broken and should be fixed Mar 15 21:11:27 kergoth_: It's conf/distro/include/multimachine.conf Mar 15 21:11:31 WORKDIR = "${TMPDIR}/work/${PACKAGE_ARCH}-${HOST_OS}/${PF}" Mar 15 21:12:04 aha Mar 15 21:14:13 pH5: the multimachine stuff could probably use some loving in general, but you could just add the HOST_VENDOR to that path for now. Mar 15 21:15:26 kergoth_: thanks, I'll do that for now. Mar 15 21:15:39 no problem Mar 15 21:18:55 koen: who killed apache on ewi? Mar 15 21:20:34 hrw: apt-get upgrade Mar 15 21:21:43 ah. ok Mar 15 21:23:23 it should be up again Mar 15 21:24:20 it is Mar 15 21:29:08 03pH5 07org.oe.dev * r1c908b4c... 10/conf/distro/angstrom.conf: split TARGET_VENDOR (-none) from TARGET_OS (linux-gnueabi) Mar 15 21:29:40 hrw: how is that testing going ? :-) Mar 15 21:30:07 not yet Mar 15 21:31:02 hrw: okay, just let me know Mar 15 21:31:37 NOTE: package qpe-presenter-0708-r0: task do_build: completed Mar 15 21:31:51 someone has CFXGA card? Mar 15 21:32:02 no, but please send me one! Mar 15 21:32:20 ;) Mar 15 21:44:34 If I put something into meta-opie under "Depends" it will be built, but not (nessesarily) put into the image, correct? Mar 15 21:47:15 kergoth_, hey, i got the PPC toolchain up. thanks for your help last week. Mar 15 21:48:07 ah nice, whatd ya have to do? Mar 15 21:49:37 first, figure out what i was doing! the good news my python has really come up to speed (-; Mar 15 21:50:40 hehe Mar 15 21:50:45 then there were some problems with various stages of glibc. that just took some experiements 'till i found one that worked. found a tip by koen in an older log that helped me towards a working version. Mar 15 21:51:33 now, i just have to figure out how to compile the usb driver, and get it loaded. Mar 15 21:51:48 (thanks koen!) Mar 15 21:55:10 pH5: I'm getting errors in libmupflap: /data/build/koen/OE/build/tmp/angstrom/cross/arm-none-linux-gnueabi/include/bits/errno.h:25:26: error: linux/errno.h: No such file or directory Mar 15 21:55:50 anyway... Mar 15 21:55:52 'night all Mar 15 21:56:03 'night koen Mar 15 21:57:48 koen|sleep: I remember seeing the same thing, probably just 'fixed' it by copying the kernel headers into cross/arm-none-linux-gnueabi/include. Mar 15 22:07:56 pH5: congrats! Mar 15 22:11:20 03pH5 07org.oe.dev * r0877c14e... 10/packages/cairo/cairo_1.0.4.bb: cairo: add bugfix release 1.0.4 Mar 15 22:20:34 woglinde: hi Mar 15 22:22:04 yo zecke Mar 15 22:37:48 'night all Mar 15 22:50:42 evening Mar 15 22:51:01 hi Rene Mar 15 22:51:11 hey Marcin Mar 15 23:09:31 cu Mar 15 23:22:02 * chouimat is away: food Mar 15 23:24:18 03yuecelm 07org.oe.dev * rf066b4f3... 10/packages/ (8 files in 6 dirs): Mar 15 23:24:25 subversion-1.3.0,davfs2-0.2.8: push newest releases Mar 15 23:24:31 - neon-0.26.0 supports gnutls, but subversion and davfs2 are yet not compatible with neon-0.26.0 Mar 15 23:24:37 - add neon-0.25.5 Mar 15 23:24:39 - again, no https support Mar 15 23:31:21 03yuecelm 07org.oe.dev * r6534eb11... 10/packages/davfs2/ (files/no-neon-version-test.patch davfs2_0.2.8.bb): davfs2: forgot to remove DEFAULT_PREFERENCE line, remove unused patch Mar 15 23:38:54 hrw|gone: when you get back, ping me, I solved my opie-embeddedconsole problem, but I still have a konq problem. Mar 15 23:41:12 hrw|gone: and ipkg-link fixed my knoq problem, sorry to waste your time. Mar 15 23:42:15 good nite guys Mar 15 23:42:49 zecke: sweet dreams Mar 15 23:56:56 * france is back (gone 11:17:42) Mar 15 23:57:03 03yuecelm 07org.oe.dev * r680ea086... 10/conf/distro/slugos-packages.conf: Mar 15 23:57:03 conf/distro/slugos-packages.conf: add some directories Mar 15 23:57:06 - apr, didiwiki, gnutls, gpsd, patchutils, subversion, time, whois Mar 16 00:16:24 * france is away: Away Mar 16 00:32:01 olla Mar 16 00:33:19 wb Mar 16 00:33:51 * emte sends chouimat some poutine Mar 16 00:47:08 * chouimat is back. Mar 16 00:59:20 hmm Mar 16 00:59:41 i wonder if viewmtn can show date ranges ... Mar 16 01:12:45 hmm Mar 16 01:12:55 well i broke one of the servers :( Mar 16 01:13:51 oh, there we go :) Mar 16 01:18:35 03yuecelm 07org.oe.dev * r2c5e0c59... 10/packages/davfs2/ (4 files in 3 dirs): davfs2-0.2.7: reactivate removed patch and move patches to the relevant version Mar 16 01:26:54 lets see if this is it Mar 16 01:28:39 * emte wonders if 15 packages testing for "generation of binary locales" is not a bit excessive Mar 16 01:49:36 I want to cross compile for the ipaq a program (fbvnc) which is not in the OE packages list. Can it be done? How? Mar 16 01:50:01 for which distro? Mar 16 01:50:14 familiar / gpe Mar 16 01:50:26 currently you may not get far .... Mar 16 01:50:37 i am chasing a build problem Mar 16 01:51:06 but basicly you build a metafile for fbvnc and get it to behave Mar 16 01:51:27 i dont know if fb support is availible in familiar or not tho Mar 16 01:51:59 i am pretty positive that gpe doesnt use fb Mar 16 01:52:15 but i have frequently been wrong Mar 16 01:52:46 There is a kind of distribution 'pocketworkstation' that basically, allows the instalation of a full debian in a removable disk. Mar 16 01:53:08 I installed it and it works all right using fb's and vnc. Mar 16 01:53:16 yeah Mar 16 01:53:27 but its not built using OE ... Mar 16 01:53:35 However, I want to modify and compile the vnc client so that it recognizes my keyboard. Mar 16 01:54:12 so some of thier work isnt availible in OE unless someone decided to share it Mar 16 01:54:12 So, what would be an appropriate tool? Mar 16 01:54:47 well there are a few options Mar 16 01:55:06 but i am not sure what the best one to suggest would be Mar 16 01:55:26 what is the reason you want a fb vnc client for? Mar 16 01:55:50 there are non-fb vnc clients already availible Mar 16 01:57:06 I did like the hendling of the mouse (three buttons), the fact that the virtual screen is much larger than the physical screen Mar 16 01:57:27 and tha possibility of rotating and rescaling in fbvnc. Mar 16 01:57:42 But if there is a good alternative, I would be glad to try it. Mar 16 01:58:15 i am not sure any of the features you described are specific to fbvnc Mar 16 01:58:27 I don't know either... Mar 16 01:58:38 those all sound liek x-server settings Mar 16 01:58:41 like* Mar 16 01:59:09 except the mouse one Mar 16 01:59:31 which confuses me since i dont know of any ipaqs taht have mice ... much less threebutton ones Mar 16 02:00:19 The mouse is emulated with the ipaq buttons and the stylus. Mar 16 02:00:53 It also has a transparent keyboard overlay. Mar 16 02:00:53 ah, must be a custom rebinding of the buttons Mar 16 02:01:02 Yeah. Mar 16 02:01:47 transparent keyboard ... sounds like maemo Mar 16 02:01:52 Anyway, what I actually want is the possibility of installing a full debian. I tried Intimate, but it failed. Mar 16 02:01:59 What's maemo? Mar 16 02:02:24 if you want to install debian then why dont you? Mar 16 02:02:35 your in the wrong place for taht tho Mar 16 02:02:58 maemo is a project for embedded systems Mar 16 02:03:32 its centrealized around the nokia770 Mar 16 02:04:02 wlm, i think your looking for this: http://www.emdebian.org/ Mar 16 02:04:47 Well, I wanted to know if I can use OE to compile. That's why I came. I don't know how to install debian on an ipaq, other than following the instructions I found for Intimate and for PocketWorkstation. I'll try embedian though. Thanks. Mar 16 02:05:14 well, the distros in OE tend to be more optimised for the hardware Mar 16 02:05:34 i cannot imagine trying to fit debian on an ipaq Mar 16 02:05:50 And by the way, I still would like to know what would be the recommended way to compile an arbitrary program for familiar. Mar 16 02:06:33 I have a 5Gb removable disk in my ipaq... Mar 16 02:13:41 emte: Thanks... Mar 16 02:35:53 man ... i've a headache ... not sure if that makes my train of thought better or worse ... Mar 16 03:46:27 nice, sylpheed-2.2.2 builds against gnutls Mar 16 03:49:02 tahts good Mar 16 03:49:30 i'll have to remember to try it if i get familiar fixed Mar 16 03:51:43 ouch, heh Mar 16 04:50:59 it is getting better tho Mar 16 04:51:04 ive got to: Mar 16 04:51:20 | ../sysdeps/unix/sysv/linux/bind.S: Assembler messages: Mar 16 04:51:20 | ../sysdeps/unix/sysv/linux/bind.S:5: Error: symbol `__bind' is already defined Mar 16 04:51:36 so that is a wee bit of an improvement Mar 16 04:51:53 firefox-1.5 bombed out on me, oji stuffs Mar 16 04:53:13 hmm Mar 16 04:58:03 lcglue.cpp:301: error: no matching function for call to ânsIJVMPlugin::UnwrapJavaWrapper(JNIEnv*&, _jobject*&, lcjsobject*)â Mar 16 04:58:06 ../../../dist/include/oji/nsIJVMPlugin.h:120: note: candidates are: virtual nsresult nsIJVMPlugin::UnwrapJavaWrapper(JNIEnv*, _jobject*, jint*) Mar 16 04:58:38 so um, i take it that jint* != lcjsobject* Mar 16 05:00:25 tahts not what it says Mar 16 05:00:34 no matching function Mar 16 05:01:26 03jnc 07org.oe.dev * r5167cde2... 10/packages/wpa-supplicant/wpa-supplicant_0.5.1.bb: Make wpa-supplicant 0.5.1 bb build again. Mar 16 05:01:33 ânsIJVMPlugin != nsIJVMPlugin Mar 16 05:02:16 â somehow got inserted or renamed i guess Mar 16 05:02:24 if they are supposed to be the same Mar 16 05:03:23 brb Mar 16 05:34:45 Luke-Jr: ping Mar 16 05:35:10 Luke-Jr: i noticed you've got some pastebin action going on, that may be relevent to firefox-1.5 not building here Mar 16 05:35:18 Luke-Jr: was there a fix you know of? Mar 16 05:44:49 jnc: dunno Mar 16 05:44:57 jnc: Firefox sucks :) Mar 16 05:45:12 jnc: my solution (IIRC) was to use a better browser as soon as one was available Mar 16 05:45:39 though I've noticed gpe-mini-browser is a bit buggy :\ Mar 16 05:45:47 but hey, it has a better HTML engine Mar 16 06:34:33 Luke-Jr: okie Mar 16 06:35:34 moo Mar 16 06:35:53 i wish i could start later chasing this .... Mar 16 06:36:22 but i guess that is what testing is all about Mar 16 06:38:39 * W8TVI plays with his new toy Mar 16 06:38:50 * W8TVI wonders why it doesn't like his SD card... Mar 16 06:39:04 (Garmin iQue 3600) Mar 16 06:39:17 nice toy Mar 16 06:39:32 got it for $111 Mar 16 06:39:40 really? Mar 16 06:39:44 where from? Mar 16 06:39:45 close out at officemax Mar 16 06:39:57 oh :( Mar 16 06:40:07 i dont think thats close to me then Mar 16 06:40:11 they are closing our local officemax Mar 16 06:40:48 they have anymore Mar 16 06:40:50 ? Mar 16 06:40:53 but I think they gave me one of the returned ones Mar 16 06:40:58 yeah they did Mar 16 06:41:31 the stupid power jack pin broke off Mar 16 06:41:36 on the pda Mar 16 06:41:58 at least it has the palm universal connector and you can charge with that Mar 16 06:42:05 ... Mar 16 06:43:10 I never expected that they'd pawn off thier junk that I'm sure they wrote off already as "new" but on close out. Mar 16 06:44:46 i wonder if its worth it to risk asking you to pick me one up at that price ... Mar 16 06:44:54 heh Mar 16 06:45:04 I would look in the box first. Mar 16 06:45:26 emte: heres mine: http://www.flickr.com/photos/w8tvi/113142736/ Mar 16 06:47:33 hrm Mar 16 06:47:48 there must be something on the card that it doesn't like Mar 16 06:48:13 emte: I know, just buy my another one of these :) Mar 16 07:00:27 well ... i narrowed it down to after 1pm on Feb11 Mar 16 07:23:14 jnc, i dont know if you will get ahold of schurig Mar 16 07:23:24 from the mailing list anyway Mar 16 07:24:15 he or his company havent been using Bitbake for a while i belive Mar 16 07:24:30 but he does stop in once every few weeks/months Mar 16 07:25:06 or wait , i should correct that Mar 16 07:25:22 they do use bitbake but the dont/wont use monotone Mar 16 07:25:25 they* Mar 16 07:25:50 i think thats how it goes Mar 16 07:43:31 [lala_]: ping Mar 16 07:44:31 he hasnt said much tonight Mar 16 08:13:48 yo Mar 16 08:28:11 morning Mar 16 08:29:13 morn Mar 16 08:30:36 good morning all Mar 16 08:31:03 morn Mar 16 08:34:49 mmm ... time to start getting ready for bed i think ... class in 5hrs Mar 16 08:37:57 cu emte Mar 16 08:39:40 yup. gnight Mar 16 08:53:07 hi mithro Mar 16 08:53:20 morning Mar 16 08:53:46 yo :) Mar 16 08:54:11 hey XorA & mithro Mar 16 08:54:16 hey zecke Mar 16 08:54:30 hey koen Mar 16 08:54:32 someone know how to setup OE for the i.mx21 ? Mar 16 08:54:32 how goes life? Mar 16 08:56:15 hi people Mar 16 08:56:20 got a question about EABI Mar 16 08:56:24 hey ArjanS Mar 16 08:56:30 mithro: up and down :) Mar 16 08:56:41 mithro: I have struggled with ArgoUML and Umbrello Mar 16 08:56:41 Genesis: mreimer might know, he has been doing i.mx stuff recently Mar 16 08:56:48 ho Mar 16 08:56:50 hi ArjanS Mar 16 08:56:50 does an EABI-built image also need an EABI-built kernel? Mar 16 08:56:57 ArjanS: yes Mar 16 08:56:59 mithro: after a couple of OOMs I finally was done Mar 16 08:57:08 ArjanS: the syscall numbers changed Mar 16 08:57:13 because embedix of lineo seems to be not free Mar 16 08:57:14 and does the kernel need much patching for that? Mar 16 08:57:19 mithro: anyway this machine had only 2GB of RAM+swap Mar 16 08:57:37 ArjanS: 2.6.16-rc5 and beyond need no patches Mar 16 08:57:46 vanilla or hh? Mar 16 08:57:55 vanilla Mar 16 08:57:57 koen: when is EABI coming to StrongARM? Mar 16 08:58:01 great! Mar 16 08:58:08 koen: thanks for the info Mar 16 08:58:16 ArjanS: as for hh.org: http://handhelds.org/hypermail/kernel-discuss/14/1472.html Mar 16 08:58:29 zecke: you'd have to ask pbrook Mar 16 08:59:28 pbrook: when is EABI coming for StrongARM? Mar 16 08:59:53 zecke: try #debian-arm :) Mar 16 09:00:04 #debian-ar: when is EABI... Mar 16 09:00:11 heh Mar 16 09:01:23 hmm I either help this project, or work on the test suite Mar 16 09:01:39 * zecke has heard rumors that the C++ bitbake parser is not too long away Mar 16 09:01:51 C++? Mar 16 09:02:17 mithro: yes, it is called parser_c to iritate Mar 16 09:02:27 zecke: wow, it uses less then rational rose :P Mar 16 09:02:42 zecke: i've got the beginnings of the wrapper on my HD Mar 16 09:02:43 zecke: not bitsoufle ? Mar 16 09:03:02 koen: well, I prefer eclair Mar 16 09:03:59 I hope we can kill the cache with the new parser :0 Mar 16 09:04:16 and sto00opid bugs Mar 16 09:05:28 koen: I think it has to do, with immediate assignments, already evaluated structures... Mar 16 09:05:38 zecke: and invariants? Mar 16 09:57:01 morning all Mar 16 09:57:39 hi dirk Mar 16 09:57:43 ~openembedded Mar 16 09:57:44 rumour has it, openembedded is a flexible build tool based on Gentoo's package manager (portage), and a large upstream package repository and build classes. It facilitates building embedded Linux distributions and packages for any arbitrary target platform. Mar 16 09:57:52 hey Marcin Mar 16 09:59:03 hmm. we need to update that definition a bit Mar 16 10:22:15 good morning Mar 16 10:28:37 hi Bernardo Mar 16 10:29:11 hey Bernardo **** ENDING LOGGING AT Thu Mar 16 10:59:56 2006