**** BEGIN LOGGING AT Wed Jul 16 02:59:56 2008 Jul 16 04:50:54 03khem 07org.oe.dev * r641a15c4... 10/ (4 files in 2 dirs): Fix rsync source path. Jul 16 04:51:07 * * OE Bug 4431 has been RESOLVED (FIXED) by raj.khem(AT)gmail.com Jul 16 04:51:08 * * rsync package: do_fech failed Jul 16 04:51:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4431 Jul 16 06:41:58 morning all Jul 16 07:13:25 RP: hi. welcome back. i'm too exhausted now to discuss, but we really need to get this thing going. I'm tired of the blocade. I will start working on the commit mails and the ciabot notification. can you take care about some other points? Jul 16 07:14:40 elo Jul 16 07:22:26 mickeyl: Yes, we need to move forwards Jul 16 07:22:44 hi hrw Jul 16 07:25:46 mickeyl: I'm working up towards further emails on the git subject Jul 16 07:27:49 great, thanks Jul 16 07:28:04 Hi, I'm new to OE and I want to compile enigma2 that uses OE for the sh4 arch. Jul 16 07:28:16 I've already downloaded and compiled it using a makefile that calls mtn, Jul 16 07:28:23 but the default config is to compile it for the mipsel arch and generate a jffs2 rootfs img. Jul 16 07:28:45 What I would like to do is just to compile enigma2 for the sh4 arch without generating any rootfs img because I already have a rootfs in which I just want to install enigma2. Jul 16 07:29:07 I'm adding a new /conf/machine/*.conf file for my arch and cross toolchain, but it's not clear what else I need to do, do you have any ideas/hints? Jul 16 07:30:20 create a your own *image.bb recipe, and add all the software you need Jul 16 07:32:01 mickeyl: mail sent Jul 16 07:34:59 Can I tell the *image.bb to compile it and install it in my rootfs similar to a 'make install'? Jul 16 07:35:57 I'm not interested at the moment in creating any ipk/rpm package Jul 16 07:37:37 thanks Jul 16 07:53:07 * * OE Bug 4433 has been created by shawn.t.rutledge(AT)gmail.com Jul 16 07:53:09 * * New recipe - Chicken Scheme 3.3.0 Jul 16 07:53:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4433 Jul 16 07:55:11 03mickeyl 07org.oe.dev * rc523797d... 10/ (1 packages/python/python-evas_cvs.bb): python-evas cvs quickhack to make it work with newer pyrex Jul 16 07:55:16 03mickeyl 07org.oe.dev * rbcdca45c... 10/ (3 files in 3 dirs): Jul 16 07:55:16 python 2.5.2 add parameter to run python in unoptimized bytecode mode Jul 16 07:55:16 Thanks to the default-is-optimized patch, this python was running in optimized Jul 16 07:55:16 mode all the time, there was no way to run it unoptimized. However this Jul 16 07:55:16 is necessary at times (e.g. if you want to run code that contains 'assert' Jul 16 07:55:17 statements, since optimized bytecode strips these out), so I added a new Jul 16 07:55:19 command line parameter 'N' to python that allows you to run it unoptimized. Jul 16 07:57:30 i have one question, it's easy to add support or make more thant one rootfs only addin the IMAGE_FSTYPE ?="ext2 jffs2" ,for example? Jul 16 07:58:12 yeap Jul 16 08:06:30 hello there! Jul 16 08:06:57 i just wrote a mail to the devel list: http://lists.linuxtogo.org/pipermail/openembedded-devel/2008-July/005751.html Jul 16 08:07:15 so if there are any quick to answer questions - feel free to ask here ;) Jul 16 08:08:56 but, in some image receipes, i have extrange genext2fs errors i thing that is a double / but i can not find any clue with grep (dependency nightmare ;)) Jul 16 08:14:51 03mickeyl 07org.oe.dev * r8df223f2... 10/ (1 conf/distro/include/preferred-om-2008-versions.inc): preferred-om-2008-versions: bump pyrex Jul 16 08:14:56 03mickeyl 07org.oe.dev * r2b803ac7... 10/ (1 conf/distro/include/moko-autorev.inc): moko-autorev.inc: remove duplicate entry Jul 16 08:15:01 03mickeyl 07org.oe.dev * r2a236f2f... 10/ (1 conf/distro/include/sane-srcdates.inc): sane-srcdates: bump EFL srcdate Jul 16 08:15:07 03mickeyl 07org.oe.dev * r0ea1a16b... 10/ (16 files in 7 dirs): remove deprecated freesmartphone.org recipes Jul 16 08:15:12 03mickeyl 07org.oe.dev * rd4251a07... 10/ (44 files in 12 dirs): remove deprecated openmoko stuff, it's no longer buildable anyways Jul 16 08:15:34 CLEANUUUUUUUUP Jul 16 08:16:22 RP: Would you look at my toolchain build sequence patch I just posted ? Jul 16 08:16:52 to me it looked good, but i'm no toolchain expert Jul 16 08:17:39 mickeyl: thanks Jul 16 08:20:45 khem: so the plan is to make uclibc more like glibc? Jul 16 08:21:09 khem: Since we have the three gcc steps with that but not with uclibc? Jul 16 08:22:39 and combine glibc-initial and glibc-intermediate Jul 16 08:23:00 building toolchains suxx Jul 16 08:23:34 gcc-initial glibc-intermediate gcc-cross glibc gcc just to get really proper libgcc1 ;( Jul 16 08:24:31 RP: yes in fact same build sequence for all Jul 16 08:24:59 even with uclibc there should be 3 steps Jul 16 08:25:38 hrw: Unfortunately we need to solve chicken egg problems in toolchain Jul 16 08:25:49 josch|nsn: I am writing a reply to your mail now Jul 16 08:25:58 Laibsch: thx :) Jul 16 08:26:01 exec summary: I've 99% in place Jul 16 08:26:07 o0 Jul 16 08:26:07 not lzma, though Jul 16 08:26:10 :( Jul 16 08:26:19 It's not what you want Jul 16 08:26:52 RP: look at packages/uclibc/uclibc.inc diff it has the DEPEND on gcc-intermediate Jul 16 08:26:58 josch|nsn: Some of my stuff is already committed in the repo Jul 16 08:27:00 khem: What you say in the email seems to make a lot of sense. I'd like to try testing it Jul 16 08:27:14 josch|nsn: for about half a year or so Jul 16 08:27:19 Laibsch: oh - i must have missed the squashfs stuff then Jul 16 08:27:23 RP: thanks that would really help Jul 16 08:27:29 josch|nsn: google for my nick and offline wikipedia Jul 16 08:27:49 khem: I will try and do that soon although I don't know how I'm going to find the time :/ Jul 16 08:27:52 josch|nsn: again, I tried squashfs, but it is not the way to go IMHO Jul 16 08:27:58 RP: I have already tested it for most common toolchain combos Jul 16 08:28:25 RP: you need a 25 hour day :) Jul 16 08:28:42 josch|nsn: "openembedded offline wikipedia" would have given you all you needed Jul 16 08:28:42 khem: indeed :) Jul 16 08:28:49 I will keep testing weird toolchain combos and see if i can break something Jul 16 08:28:50 Laibsch: is there a quick description of your work anywhere? Jul 16 08:28:51 josch|nsn: we even have an open bug for it Jul 16 08:28:58 josch|nsn: google Jul 16 08:29:24 the git repo is only used for checkouts or even for commits as of now ? Jul 16 08:29:57 khem: checkouts Jul 16 08:30:06 RP: thx Jul 16 08:31:13 good night all Jul 16 08:31:31 * khem goes to bed Jul 16 08:32:47 Laibsch: okay but my request is not so much about offline wikipedia as about squashfs-lzma Jul 16 08:35:41 josch|nsn: do you want a recipe or do you want a solution? Jul 16 08:35:50 * Laibsch is usually looking for solutions Jul 16 08:36:23 03koen 07org.oe.dev * rd164ddcc... 10/ (5 files in 3 dirs): Jul 16 08:36:23 dsplink: add glue to hook in TI dspbios and TI c64x compiler trees into the build Jul 16 08:36:23 * The user needs to obtain dspbios and cgtools on its own and override the TI*DIR vars in local.conf to point to the correct location, default is /OE/TI Jul 16 08:36:38 josch|nsn: BTW, google for it and I think you will find that the lzma/squashfs stuff has been done as well Jul 16 08:41:21 josch|nsn: http://bugs.openembedded.net/show_bug.cgi?id=3711#c2 Jul 16 08:41:35 It's all right there in the open Jul 16 08:41:59 research skillz Jul 16 08:42:07 't was the first hit in google, too Jul 16 08:42:48 200% increase, IOW triple size Jul 16 08:44:23 good morning Jul 16 08:45:24 good morning, florian Jul 16 08:48:57 Laibsch: i already read the bug Jul 16 08:49:26 Laibsch: i also googled for squashfs lzma in openembedded and asked some openmoko guys but didnt find any existing packages Jul 16 08:49:45 and what do you mean by the size increase? Jul 16 09:04:10 03koen 07org.oe.dev * rdaafad01... 10/ (4 files in 3 dirs): dsplink: remove some trailing spaces Jul 16 09:06:05 Laibsch: just ping if you are around again Jul 16 09:08:58 hrm... looks like having TMPDIR on nfs was a bad idea Jul 16 09:09:18 florian: it's not a great idea for speed, but it ought to work fine Jul 16 09:09:31 all my home directories are on nfs and I fairly often have TMPDIR in there Jul 16 09:10:36 pb__: It works but it is terribly slow... I guess there is some room for optimisation with a better nfs configuration Jul 16 09:10:53 right, could be Jul 16 09:11:08 I've never found it to be unbearably slow. it's measurably slower than local disk but not so much so that it usually upsets me. Jul 16 09:12:44 pb__: sounds like I sould take a look at the configuration right now... it is extremely slow. Jul 16 09:13:42 :-( Jul 16 09:14:18 florian: It it mounted sync or async? Jul 16 09:14:26 Is it... Jul 16 09:17:40 RP: Whatever the default setting might be - I guess async and larger rsize/wsize might give a lot more performance. Jul 16 09:20:07 * * OE Bug 4434 has been created by janprunk(AT)gmail.com Jul 16 09:20:09 * * Replacing 2007 Angstrom feeds with 2008 causes dbus to crash Jul 16 09:20:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4434 Jul 16 09:20:28 florian: yah, the default settings are very conservative Jul 16 09:24:32 RP: Thanks - that was the right question... async reduces the load of the server as well. Jul 16 09:25:33 03koen 07org.oe.dev * r027df8bf... 10/ (1 packages/linux/linux-omap2_git.bb): linux omap2 git: bump to 2.6.26 Jul 16 09:28:47 RP: Do you have an idea about #4428? Jul 16 09:30:06 hi all Jul 16 09:30:50 hi gunny Jul 16 09:31:41 hi florian Jul 16 09:31:51 got a question Jul 16 09:31:53 does monotone repository have a web interface for simple browsing? Jul 16 09:32:46 i remember a file browser available from the prev ver of www.openembedded.org Jul 16 09:33:18 yes right Jul 16 09:33:22 hrmm Jul 16 09:33:34 * florian wonders where it is now Jul 16 09:34:03 Laibsch: Where does the filebrowser hide now? Jul 16 09:34:53 I have no idea Jul 16 09:35:02 I did not rework the website Jul 16 09:35:18 mickey didn't either Jul 16 09:35:24 we both don't know who did Jul 16 09:35:42 florian: use the git file browser Jul 16 09:36:10 gunny: http://wiki.openembedded.net has a link Jul 16 09:36:16 gunny: http://gitweb.openembedded.net/ Jul 16 09:36:19 heh Jul 16 09:36:38 thanks, florian Jul 16 09:36:38 Laibsch: I just saw your name in the changelog :) Jul 16 09:36:59 well, not related to the website change, or was it? Jul 16 09:37:04 ltg changelog? Jul 16 09:37:29 I have not touched ltg for at least 3 months, probably more Jul 16 09:37:30 thanks Laibsch Jul 16 09:37:33 thanks florian Jul 16 09:37:49 you're welcome, gunny Jul 16 09:38:32 florian: offhand, no idea about 4428, :( Jul 16 09:38:58 * Laibsch reckons that the ltg apache would still serve requests if the DNS were to resolve it Jul 16 09:39:37 Laibsch: I guess its just a matter of a link on the new website. Jul 16 09:39:40 ~curse shoddy communications Jul 16 09:39:41 May you be reincarnated as a Windows XP administrator, shoddy communications ! Jul 16 09:40:13 on gitweb.openem... am I browsing a GIT repo of the oe tree? Jul 16 09:40:28 RP: ok, thanks... will take a look later Jul 16 09:40:46 XorA: there is none, plain and simple ;-) all entries in Changelog on amethyst were from me last time I looked. I doubt I was the only one making changes Jul 16 09:40:56 gunny: yes Jul 16 09:41:09 Laibsch: are you talking the www.openembedded.org->wiki switch Jul 16 09:41:11 same thing, different software Jul 16 09:41:19 XorA: yes Jul 16 09:41:24 Laibsch: was koen Jul 16 09:41:30 OK Jul 16 09:41:34 is it an alternative to the monotone repo? Jul 16 09:41:35 mickey|lunch: knows this Jul 16 09:41:44 I never expected www. to be replaced by or redirected to wiki. Jul 16 09:41:46 or should do if he reads email :-) Jul 16 09:41:52 o.... Jul 16 09:42:09 I never noticed before that http://openembedded.org/ => wiki Jul 16 09:42:21 I noticed a few days ago Jul 16 09:42:26 hrw: you not reading email either Jul 16 09:42:38 few days ago I was on guadec where mail ports were blocked Jul 16 09:42:46 ~fish hrw Jul 16 09:42:47 * ibot slaps hrw around with a large trout Jul 16 09:42:52 hrw: and then placed a few links into the wiki to make sure that git and tinderbox were at least reachable Jul 16 09:42:54 your supposed to be an uber hacker :-) Jul 16 09:43:25 those were the services I was aware of Jul 16 09:43:35 hrw: They weren't blocked at the hotel ;-) Jul 16 09:43:35 I asked for other services to link to here but got no response Jul 16 09:46:33 RP: where I mostly slept ;) Jul 16 09:46:57 RP: I looked at OH mails mostly in hotel and how home machine behave Jul 16 09:51:01 RP: did you guys have fun in istanbul? Jul 16 09:51:43 XorA: Yes, was an interesting change of scenery :) Jul 16 09:51:59 RP: cool, Im annoyed I missed it Jul 16 09:55:06 * * OE Bug 4435 has been created by ospite(AT)studenti.unina.it Jul 16 09:55:08 * * openmoko-appmanager2_svn.bb fails in do_configure Jul 16 09:55:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4435 Jul 16 10:27:25 Laibsch: it seems you dont want to talk about this matter? Jul 16 10:33:22 mickey|lunch: Your mail to Mike did not end up in his mailbox I guess ;) Jul 16 11:27:14 03koen 07org.oe.dev * r929db1be... 10/ (1 packages/vlc/vlc-davinci_0.8.6h.bb): vlc-davince: update srcrev, use vlcs compile system (yuck!), fix description Jul 16 11:27:19 03koen 07org.oe.dev * reb0e2717... 10/ (3 files in 3 dirs): codec-engine: hook into TI xdc and dsp toolchains Jul 16 11:40:14 03koen 07org.oe.dev * r2b7b118c... 10/ (1 packages/dsplink/codec-engine_2.10.bb): codec-engine: more bruteforce sed to get rid of bogus target-prefixes Jul 16 12:07:00 Hi all Jul 16 12:07:42 I search a page in the new wiki without sucess : FirstProject Jul 16 12:07:56 it was availlable before the update Jul 16 12:08:16 http://web.archive.org/web/20070502104256/http://www.openembedded.org/wiki/FirstProject Jul 16 12:08:46 Do you have a replacement entry ? Jul 16 12:12:26 add it? Jul 16 12:18:19 it would help to have the old page still available for that Jul 16 12:18:29 is it still relevant "as is" ? Jul 16 12:18:40 then one would be able to get the mark-up for the page Jul 16 12:20:47 Laibsch: if you dont want to answer me or have no time just say so and i will remain silent but i have a couple of things i do not understand about what you said Jul 16 12:23:05 what is it? Jul 16 12:23:08 good morning Jul 16 12:23:30 Laibsch: well first you said that there already is squashfs+lzma in openembedded? Jul 16 12:23:39 I said I think it is Jul 16 12:23:49 okay - there is not ;) Jul 16 12:24:04 And given my impression you did not do your research on the other subjects, either, I am not willing to do that for you Jul 16 12:24:11 josch|nsn: I am still not convinced Jul 16 12:24:21 there is a lot of squashfs and lzma stuff Jul 16 12:24:32 I have another wiki page that I cannot find : PapersAndPresentations Jul 16 12:24:32 So I think there will also be a kernel module somewhere Jul 16 12:24:43 steliosk would be the one to talk to, I think Jul 16 12:25:18 noglitch: We didn't do a complete transfer Jul 16 12:25:19 Laibsch: as i said this is not only for my project but in my opinion would be a great addition to openembedded in a whole Jul 16 12:25:31 noglitch: people added what they felt was important Jul 16 12:25:44 josch|nsn: and I said I think it's there Jul 16 12:25:53 If it's not, then it's not Jul 16 12:25:58 Laibsch: sry didnt want to offend you Jul 16 12:26:00 I cannot help you write it Jul 16 12:26:06 don't worry Jul 16 12:26:33 can you also tell me what you mean about the 200% size increase you observed? Jul 16 12:26:51 squashfs module is in linux-rp 2.6.24 2.6.23 Jul 16 12:27:03 patch probably just needs updated for newer ones Jul 16 12:27:04 josch|nsn: There you go Jul 16 12:27:04 XorA: but without lzma comnpression Jul 16 12:27:10 and its in linux-wrt as well Jul 16 12:27:13 with lzma Jul 16 12:27:23 XorA: thanks Jul 16 12:27:25 and linux-magicbox Jul 16 12:27:36 XorA: thanks i was unable to find anything so far Jul 16 12:27:44 then there has to be a bb for it Jul 16 12:27:53 josch|nsn: original size was 700MB, recompressed as squashfs filesystem was 2G Jul 16 12:27:59 find packages -name "*squash*" Jul 16 12:28:00 clearly stated in the bug report Jul 16 12:28:13 I even sent you the *exact* URL Jul 16 12:28:26 Laibsch: i read all that in this is why i asked Jul 16 12:28:38 and that *does* make me mad, that after you did not research this yourself, you fail to read Jul 16 12:28:54 Laibsch: i also found your bugreport via google before you posted the link Jul 16 12:28:59 what is unclear about the message there? Jul 16 12:29:04 good Jul 16 12:29:06 well done Jul 16 12:29:09 what is unclear about the message there? Jul 16 12:29:41 i assume 700mb of the wikipedia xml dump? Jul 16 12:29:51 but which dump? Jul 16 12:29:55 does it really matter? Jul 16 12:30:03 I think it was the html dump Jul 16 12:30:07 well i did compress great with squashfs-lzma Jul 16 12:30:17 it was even better than the original Jul 16 12:30:21 if you think the result can be improved with a different file, go ahead and try it Jul 16 12:30:30 as i said i have dewiki in 1.3GB with ALL formattting Jul 16 12:30:37 Laibsch: ok, thanks Jul 16 12:30:41 i not only tried it - i did it Jul 16 12:30:46 well, I have my doubts you have the complete wikipedia with just 1.3G Jul 16 12:30:56 or you don't have it as a filesystem Jul 16 12:30:58 this is what makes me wonder about your results Jul 16 12:31:05 if you found something, I overlooked, cool Jul 16 12:31:10 Laibsch: i have - you want to download and have a look yourself? Jul 16 12:31:17 make a comment in the bug report and I'll look at it when I have time Jul 16 12:31:34 its the complete dewiki with all formatting, linking, redirect, categories... Jul 16 12:32:01 and everything as a filesystem - one can of course use this uncompressed too Jul 16 12:32:01 well, I really dont'have time right now Jul 16 12:32:07 no problem but Jul 16 12:32:09 but since the offer is so sweet ;-) Jul 16 12:32:17 this was why i was wondering about your results Jul 16 12:32:24 as i experienced the opposite Jul 16 12:32:26 OK, let's try Jul 16 12:32:51 What did you use for downloading the orig file? Jul 16 12:32:54 this torrent: http://mokopedia.mister-muffin.de/raw-attachment/wiki/WikiStart/mokopedia.lzma.torrent Jul 16 12:33:08 the original is the wikipedia xml sump Jul 16 12:33:11 *dump Jul 16 12:33:30 i imported it to a local mediawiki install and dumped with an optimized script Jul 16 12:33:42 uncomrpessed size of dewiki is 6.8GB Jul 16 12:33:50 You lied Jul 16 12:33:51 ;-) Jul 16 12:33:55 It's 1,4G Jul 16 12:33:57 ;-) Jul 16 12:34:40 i'm currently dumping enwiki - will take two days now Jul 16 12:35:21 so how do you do updates for these portable wikis, is there some sort of diff tool? Jul 16 12:36:57 as i was told by the wikimedia-tech guys diff is useless Jul 16 12:37:00 I guess somebody has to do the whole thing over again Jul 16 12:37:09 yes, I would assume so Jul 16 12:37:17 One person does it Jul 16 12:37:19 in half a year a diff would be nearly as bug a sthe complete dump Jul 16 12:37:24 and then it is distributed via torrent Jul 16 12:37:33 sounds OK to me Jul 16 12:37:39 and the update itself would be too ressource intensive Jul 16 12:37:53 you would have to update 7mill files/articles Jul 16 12:38:34 also: its an encyclopedia so no uptodate info is needed Jul 16 12:38:56 i would also be happy with a two year old encyclopedia Jul 16 12:39:57 me too Jul 16 12:40:09 josch|nsn: are you seeding? Jul 16 12:40:13 I don't see anybody Jul 16 12:40:19 oups Jul 16 12:40:25 let me recruit a few more seeders Jul 16 12:41:36 yes Jul 16 12:41:52 I'd be happy to seed loooooooong time afterwards Jul 16 12:41:58 if it does what you promise Jul 16 12:42:07 nice Jul 16 12:42:16 BTW, I'd like to see the Japanese dump even more than the German one Jul 16 12:42:26 the only problem now is the squashfs-lzma thingy as it is painfull to buld without an ipk Jul 16 12:42:38 Plus, what is most interesting are the dump scripts and procedures Jul 16 12:42:44 those need to be documente Jul 16 12:42:45 d Jul 16 12:42:46 Laibsch: i have instructions how to build a dump here: http://mokopedia.mister-muffin.de/wiki/CreatingTheDump Jul 16 12:43:00 but one step after the other, need to find out if really works Jul 16 12:43:02 everything is there Jul 16 12:43:09 josch|nsn: cool, thanks Jul 16 12:44:42 josch|nsn: Why do you have to dump yourself? Jul 16 12:44:52 Are you aware of download.wikimedia.org? Jul 16 12:45:08 i get my dumps from there Jul 16 12:45:22 as you can read i download the xml dumps there Jul 16 12:45:23 but they have xml and html dumps, too Jul 16 12:45:29 OK Jul 16 12:45:33 the html dumps suck for various reasons Jul 16 12:45:57 Your end-product is xml or html or something else? Jul 16 12:46:01 html Jul 16 12:46:09 because on an embedded device Jul 16 12:46:15 parsing is not an option Jul 16 12:46:21 yes Jul 16 12:46:38 we still have problems even with a simple browser ;-) Jul 16 12:47:44 i only supply a http server - the browser is the user's choice as there are so many differnt screen sizes etc Jul 16 12:48:15 this makes it also possible to distribute the data over network Jul 16 12:48:42 for examle two openmoko users meet, create a mesh network and one reads wikipedia from the other's device Jul 16 12:50:14 Laibsch: i got three other people to seed it again but i was told that there are some temporary problems with the tracker (denis.stalker.h3q.com) Jul 16 12:55:57 cool Jul 16 12:56:02 I'm maxing out Jul 16 12:56:42 josch|nsn: How do you index and make it searchable? Jul 16 12:56:48 with sphinx Jul 16 12:57:01 URL? Jul 16 12:57:09 http://www.sphinxsearch.com/ Jul 16 12:57:37 i got the dev to write fixes for armv4t so now it works on my neo1973 Jul 16 12:59:54 nice Jul 16 12:59:57 good work Jul 16 13:00:06 why did you choose sphinx? Jul 16 13:00:27 for four character words the index is 1/5 of the original size Jul 16 13:00:48 and on the desktop it searches 6.8GB wikipedia fulltext in 0.02 secs Jul 16 13:01:06 well, size and speed is nice Jul 16 13:01:16 as long as it does what it is supposed to dod Jul 16 13:01:17 do Jul 16 13:01:19 for title search i use simple sqlite searching Jul 16 13:01:46 Laibsch: the results are even sorted by number of occurences etc Jul 16 13:01:49 can it do UTF-8, CJKV (asian languages)? Jul 16 13:02:06 hrm... never even tried german umlauts Jul 16 13:02:06 How do you interface to it? Jul 16 13:02:14 Laibsch: there is a python module Jul 16 13:02:26 right now you use the command line? Jul 16 13:02:47 yepp - but including it into my server is a no-brainer Jul 16 13:03:26 great Jul 16 13:03:37 looks like we got everything in place Jul 16 13:03:56 "my server" = your physical server Jul 16 13:04:15 "my server" = some software server to be included in OE? Jul 16 13:05:00 what do you mean? Jul 16 13:05:20 ah Jul 16 13:05:38 "my server" = the server application that serves the html to the browser Jul 16 13:05:50 and allows searching and multiple languages and stuff Jul 16 13:06:02 so one has not to cram around in the filesystem Jul 16 13:06:13 and gets nice language dependent css Jul 16 13:07:47 How do you get OE to include the kernel modules in an image instead of making a separate modules image? Jul 16 13:08:57 all modules? Jul 16 13:09:05 install "kernel-modules" package Jul 16 13:10:54 hrw: Ah ok. Thank you :) Jul 16 13:12:56 hrw: Is there another name for it? Jul 16 13:13:08 kernel-modules and linux-modules don't exist Jul 16 13:13:52 nvm Jul 16 13:16:03 hello Jul 16 13:16:16 i've got a rather big problem with OE on my zipit z2.. Jul 16 13:16:39 when booting to oe i get this errormessage: Jul 16 13:16:40 "switch_root: bad newroot /sysroot" Jul 16 13:18:51 josch|nsn: Sounds great Jul 16 13:19:07 Where is that server package? Jul 16 13:19:28 Laibsch: no package yet - still have to write a bb for it Jul 16 13:19:45 but you can have a look at the tarball Jul 16 13:19:46 I'll gladly help if you need it Jul 16 13:19:50 yeah Jul 16 13:20:42 Got the mokopedia.lzma file Jul 16 13:21:17 Laibsch: already finished??? Jul 16 13:21:22 Laibsch: http://rabenfrost.net/mokopedia/mokopedia.tgz Jul 16 13:21:48 yes Jul 16 13:22:13 this needs the python-opendir python-netserver packages Jul 16 13:22:14 some kind soul at vodafone provided me with all the download I could handle ;-) Jul 16 13:22:42 Laibsch: ah this was emdete Jul 16 13:23:14 I've heard that name somewhere Jul 16 13:24:06 I think that must go way back to Usenet times Jul 16 13:25:06 he is now around openmoko and stuff Jul 16 13:25:12 cool Jul 16 13:26:21 Is there a way to rebuild the kernel in OE without doing a full -crebuild? Jul 16 13:30:54 -f -ccompile Jul 16 13:31:16 hrw: Ooooh. There's an -f flag Jul 16 13:33:49 josch|nsn: Looks like you're settled and you've basically done it. What else do you need? Let me know if there is any input left you'd be looking for. Jul 16 13:34:38 Laibsch: i currently organize the squashfs-lzma thingy with shoragan Jul 16 13:34:55 what he found in oe was only the squashfs-lzma-tools Jul 16 13:34:59 not the kernel modules Jul 16 13:35:04 there are some patches Jul 16 13:35:09 but for very old versions Jul 16 13:36:00 josch|nsn: BTW, you are aware of http://users.softlab.ece.ntua.gr/~ttsiod/buildWikipediaOffline.html, the method, I was pursuing until now, aren'T you? Jul 16 13:36:28 03koen 07org.oe.dev * rc6784367... 10/ (1 packages/dsplink/codec-engine_2.10.bb): codec-engine: pass CFLAGS to demos Jul 16 13:36:34 XorA told you where you needed to look, though, didn't he? Jul 16 13:36:52 Laibsch: i am Jul 16 13:37:04 piroko: there is --help flag too Jul 16 13:37:31 Laibsch: i still fail to find modules for 2.6.24 and squashfs 3.3 and lzma457 Jul 16 13:37:37 hrw: Yeah, I missed the -f flag. My apologies Jul 16 13:37:41 those that are there are twop years old Jul 16 13:38:44 Laibsch: do you want to help in comparing the various differnt offline wikipedias up to now? http://mokopedia.mister-muffin.de/wiki/Comparison Jul 16 13:39:25 I am not an expert on that Jul 16 13:39:39 I'd be more interested in getting it working OOTB in OE Jul 16 13:39:51 Let's find and update if necessary those modules Jul 16 13:40:05 * Laibsch finally gets of his arse to search for the stuff Jul 16 13:40:09 off Jul 16 13:40:32 Laibsch: as far as i know the method by Thanassis Tsiodras does the rendering on the device? Jul 16 13:40:49 it also does not use the mediawiki parser so some things get rendered wrong Jul 16 13:43:05 I think it takes the html dump, abuses bzip2recover for some clever indexing, but then basically presents the html files to be displayed by a browser Jul 16 13:43:13 embeddable, I think Jul 16 13:43:42 already cleaner and more supportable than the dump scripts in perl out there Jul 16 13:44:05 your solution seems to be even cleaner and more supportable Jul 16 13:44:14 let's hope it keeps its promises Jul 16 13:44:27 we will see how much i can compress enwiki Jul 16 13:44:33 this is the holy grail ^^ Jul 16 13:45:27 not for me Jul 16 13:45:33 for me it is jawiki Jul 16 13:45:35 ;-) Jul 16 13:46:00 but that is the good thing about FOSS, it supports multiple preferences and choice Jul 16 13:46:15 josch|nsn: http://rafb.net/p/eeqXb419.html Jul 16 13:46:24 that is what is there Jul 16 13:46:44 Like XorA said, we probably should try and adapt the magicbox stuff to linux-rp Jul 16 13:47:09 which magicbox stuff? Jul 16 13:47:32 /export/oe/org.openembedded.dev/packages/linux/linux-wrt-2.4.30/600-linux-squashfs-lzma Jul 16 13:47:38 Is that possibly all we need? Jul 16 13:47:59 2.4 patch for 2.6 kernel? Jul 16 13:48:01 Laibsch: most of that is squashfs-tools there are also some 2.6.19 patches which i am aware of but which are outdated Jul 16 13:48:11 hrw: oops Jul 16 13:48:13 ;-) Jul 16 13:48:16 Hehe Jul 16 13:48:20 Overlooked that Jul 16 13:48:43 we at least need squashfs 3.3 as it supports 1MB blocks - shoragan seems to do some patching atm Jul 16 13:48:56 josch|nsn: So, this means the patch has to be written and not just stolen? Jul 16 13:49:09 well, glad to hear Jul 16 13:49:35 Laibsch: http://www.squashfs-lzma.org/ Jul 16 13:50:10 hrw: i write him this address an email this morning Jul 16 13:50:20 *wrote Jul 16 13:50:41 Laibsch: the patch is written Jul 16 13:51:25 hrw: the author of the patches said that this site is not actively maintained and those patches work but are not guarantued to do so Jul 16 13:51:43 so i suggest to use the patches that he actively uses to build the slax kernel Jul 16 13:51:44 Ok, I'll take a backseat for the rest of the day then and let the experts take over Jul 16 13:51:58 Laibsch: hey this is my line! :P Jul 16 13:52:58 shoragan: you should try the patch in here: ftp://ftp.slax.org/source/slax/kernel/2.6.24.3/src-core/patches/squashfs/sqlzma.tar.bz2 Jul 16 13:53:59 shoragan: sqlzma2k-3.3.patch Jul 16 13:54:30 oh and before that of course: squashfs3.3/kernel-patches/linux-2.6.24/squashfs3.3-patch Jul 16 13:55:46 Is it possible to disable the "Synthesizing hotplug events" when udev starts up? Jul 16 13:57:41 Laibsch: as for sphinx from the doc: "supports both single-byte encodings and UTF-8;" Jul 16 13:57:55 Laibsch: the dev is russian so this seems only natural ;) Jul 16 13:59:33 Trying to speed up the boot process here... Jul 16 13:59:43 great Jul 16 14:03:29 does anyone mind helping me out with a question relating to ssp/spi? Jul 16 14:03:51 I can't figure out exactly which modules we need to enable in OE to get ssp/spi working Jul 16 14:04:10 pxa2xx_spi builds fine Jul 16 14:04:20 and loads on our system Jul 16 14:04:43 but when we try to run a custom kernel module that we wrote to use ssp_init, ssp_config, ssp_end, etc. the symbols are not found Jul 16 14:04:53 and thsoe are defined in linux/arch/arm/mach-pxa/ssp/ssp.c Jul 16 14:05:05 so I would assume that ssp.c would be loaded into pxa2xx_spi but it appears not to be so Jul 16 14:05:31 make it depend on ssp? Jul 16 14:05:45 in the kernel configuration? Jul 16 14:05:56 or in our custom module? Jul 16 14:06:44 kernel configuration of your module Jul 16 14:09:36 pxa2xx_spi shouldn't make any difference if you're not using it - the SSP API is configured separately and also used for other things (eg, audio). Jul 16 14:12:57 how does OE determine to use an older version of a package? I want it to use a newer version without specifying a PREFERS Jul 16 14:20:10 checked distro configs? Jul 16 14:21:04 hrw: Will bitbake copy the defconfig over if I just do a -f -ccompile on linux? Jul 16 14:21:10 hmm Jul 16 14:22:59 no Jul 16 14:53:06 * * OE Bug 4436 has been created by drwyrm(AT)gmail.com Jul 16 14:53:08 * * configure script of mpfr-native fails on debian-etch + some backports Jul 16 14:53:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4436 Jul 16 15:35:08 What exactly does udevtrigger do for a machine? It doesn't load any modules at all... Jul 16 15:35:55 But when I comment it out, my system just stops booting at Finished configuring packages... Jul 16 15:42:20 Ah. Udev trigger amkes all the /sys stuff... Jul 16 16:25:46 Does anyone here have a spitz handy they could do some testing on? Jul 16 16:44:51 shoragan, hey - what are your findings so far? Jul 16 16:46:43 bye Jul 16 16:51:07 03thebohemian 07org.oe.dev * rd081501e... 10/ (4 files in 2 dirs): Jul 16 16:51:07 cacao-native 0.99.2: Fixed expected md5sum. Jul 16 16:51:07 cacaoh-native 0.99.2: Fixed expected md5sum. Jul 16 16:51:07 cacao 0.99.2: Fixed expected md5sum. Jul 16 16:59:05 03mickeyl 07org.oe.dev * r6effd4b9... 10/ (3 files in 3 dirs): add fluidsynth, a software synthesizer. bb by thomasg Jul 16 17:26:26 Wait, isn't the power button on the 720 at least supposed to dim the display, lower the cpu, and kill the pcmcia slot? Jul 16 17:26:54 It did in 2.4.35 Jul 16 17:33:21 Whoops. Wrong channel. Sorry Jul 16 17:51:34 Laibsch, hehehehe - i get down even further! I just saw that i also dump all the pages of the Wikipedia namespace but they are no use in an offline wikipedia. they make up 25% of the database size :D Jul 16 17:52:18 awesome Jul 16 17:52:50 I do wonder how much that 25% translates into after-compression reduction, though Jul 16 18:00:03 hah, (NOTE: kergoth says it will take about 4GB of disk space to build an opie or gpe image for one architecture) is on the gettingstarted page Jul 16 18:00:11 something tells me it needs an update Jul 16 18:10:37 Laibsch, this is why i now will finish the enwiki dumping with this namespace, compress, remove the namespace and then compress again to compare both Jul 16 18:11:04 enwiki is appearently heavily spammed :( Jul 16 18:17:09 hmm, libstdc++6 is not working in IMAGE_INSTALL :-( Jul 16 18:17:23 wonder if the '++' is messing opkg up? Jul 16 18:19:01 actually, opkg is fine, it is bitbake that is confused Jul 16 18:32:38 woglinde disappeared Jul 16 18:49:12 03mickeyl 07org.oe.dev * rddc393a9... 10/ (1 conf/distro/include/sane-srcrevs.inc): sane-srcrevs.inc: bump openmoko-appmanager2 (fixed a configure bug), remove SRCREVS for removed entries Jul 16 18:52:06 * * OE Bug 4083 has been RESOLVED (INVALID) by Jul 16 18:52:08 * * openmoko-messages2-0.1.0+svnr3868-autobuild Jul 16 18:52:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4083 Jul 16 18:55:06 * * OE Bug 4030 has been RESOLVED (FIXED) by Jul 16 18:55:08 * * openmoko-panel-gsm-0.1.0+svnr3664-autobuild Jul 16 18:55:08 03mickeyl 07org.oe.dev * r19d4d4dc... 10/ (1 packages/openmoko-panel-plugins/openmoko-panel-gsm_svn.bb): openmoko-panel-gsm: DEPEND on libmokogsmd2 Jul 16 18:55:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4030 Jul 16 18:58:05 * * OE Bug 4435 has been RESOLVED (FIXED) by Jul 16 18:58:08 * * openmoko-appmanager2_svn.bb fails in do_configure Jul 16 18:58:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4435 Jul 16 19:08:25 * kergoth`work chuckles .. followed getting started, fresh bitbake+oe .. build fails due to fetch errors... ah.. memories Jul 16 19:09:11 which getting started? .org or .net? Jul 16 19:09:43 hmm, they are the same now :) Jul 16 19:10:59 indeed Jul 16 19:11:28 * kergoth`work still thinks oe & bitbake need major architectural changes Jul 16 19:12:28 * piroko nods heavily Jul 16 19:13:12 kergoth`work, fetch errors during build? Jul 16 19:13:18 ie source fetches? Jul 16 19:13:22 yeap Jul 16 19:13:56 heh Jul 16 19:14:11 * kergoth`work craves an apt-like OE recipe manager Jul 16 19:14:15 things move, and it takes some one doing a fresh buils to notice Jul 16 19:14:16 hi Jul 16 19:14:50 woglinde: Hey! Jul 16 19:14:58 yeah, i know. just sad when one of the "bitbake foo"'s recommended on gettingstarted fail :P Jul 16 19:15:03 doesn't do well for the image Jul 16 19:18:00 kergoth`work: pretty much describes OE right now, we're working on it :-) Jul 16 19:19:16 i'm slowly getting the feeling that OE definately wipes the floor with all of its competition right now.. but given the competition, that isn't saying much Jul 16 19:19:19 so many things so little time Jul 16 19:19:23 so there's still a long way to go :P Jul 16 19:19:26 heh Jul 16 19:19:53 kergoth`work, people inside TI are using OE now :) Jul 16 19:20:03 nice, what department? Jul 16 19:20:18 beagle I guess Jul 16 19:20:53 beagleboard guys Jul 16 19:21:04 woglinde: Any progress at all on libftdi? :( Jul 16 19:21:19 kergoth`work, I don't know what groups Jul 16 19:21:24 as in the TI names Jul 16 19:21:40 piroko about what ? Jul 16 19:21:40 piroko sorry I have not the time yet Jul 16 19:21:41 hmm, cool Jul 16 19:21:47 I hope next week Jul 16 19:22:10 gremlin[it]: I have this package (libftdi) that uses rpath ugly-ish, and I can't get openembedded to compile it for the life of me Jul 16 19:23:10 gremlin[it]: http://pohl.ececs.uc.edu/~jeremy/libftdi_0.13.bb if you wanna have a go at it! Haha Jul 16 19:23:14 piroko what libftdi do ? Jul 16 19:23:41 It's a library for accessing a usb-serial chip by ftdi Jul 16 19:23:50 jupp Jul 16 19:24:14 piroko but the serial part is already available via kernel driver ... Jul 16 19:24:23 gremlin[it]: No easy way to access it though Jul 16 19:24:29 gremlin[it]: Especially in bit-bang mode Jul 16 19:24:43 Thus, libftdi. It talks directly to the chip without a kernel driver Jul 16 19:27:57 piroko, right kernel module don't allow bit-bang ... but i found a modified kernel modules source with some new IOCTL to allow eeprom manage and bit-bang :) Jul 16 19:36:47 gremlin[it]: Is the API easy to use? Jul 16 19:36:53 libftdi is trivial Jul 16 19:38:12 piroko, from the source i see : yes ... just a couple of IOCTL to read and write bit-bang IO Jul 16 19:39:12 gremlin[it]: Huh. Do you have a link to the patches perchance? Jul 16 19:42:41 piroko "ftdi-usb-sio" of sourceforge.net Jul 16 19:43:36 gremlin[it]: This is different from the kernel's built-in ftdi_sio? Jul 16 19:43:52 piroko, yes Jul 16 19:44:06 gremlin[it]: Awesome. I'll give it a try. Thank you Jul 16 19:44:47 gremlin[it]: And I assume I should grab 1.3.1 with T J bitbang? Jul 16 19:45:42 piroko, i found it cause i need to set the bit-bang bits to reset a daughter board and start in boot-mode, sent the new firmware via serial port (via ftdi) and the reset again ... Jul 16 19:45:48 piroko, exactly ! Jul 16 19:52:17 gremlin[it]: Do you happen to have any example code which uses this driver? Jul 16 19:52:43 piroko ... no i#ll write next week ... Jul 16 19:53:25 I cam Jul 16 19:53:31 *I can't even get it to compile Jul 16 19:54:01 piroko ... oh bad .... Jul 16 19:54:08 morning Jul 16 19:54:54 piroko you know if libftdi is usable also to do normal RS-232 or only eeprom and bit-bang ? Jul 16 19:55:21 gremlin[it]: It can do both Jul 16 19:59:47 piroko, thanks i'll check what would be better for my issue ! (libftdi or patched kernel driver) Jul 16 20:02:07 gremlin[it]: libftdi is certainly easier Jul 16 20:14:03 piroko, i Hope ... but i already have software that send the firmware via normal serial port ... Jul 16 20:23:39 gremlin[it]: So you what works Jul 16 20:23:41 *use Jul 16 20:24:37 i'll see ... i start to code next week ... this week i only search for available tools ... Jul 16 20:25:08 Lots of time to spend on each task O_o Jul 16 20:30:00 hahah no no ... one day this week, one day next week ... Jul 16 20:30:34 i'm doing a massive time-division-multiplex of various projects at work :) Jul 16 20:50:23 03crofton 07org.oe.dev * re5e5a981... 10/ (1 packages/u-boot/u-boot_git.bb): u-boot_git : Update revs for davinci machines davinci EVM and SFF SDR. Jul 16 21:00:12 lumdidum Jul 16 21:03:48 have yall seen this http://youtube.com/watch?v=e-LOtKIIKcg Jul 16 21:03:52 have yall seen this http://youtube.com/watch?v=e-LOtKIIKcg Jul 16 21:03:59 haha Jul 16 21:04:43 what package would flash_eraseall be part of? would think there would be flashutils or something Jul 16 21:09:19 mtd-utils? Jul 16 21:12:37 should be mtdutils, yes Jul 16 21:12:49 hi rp Jul 16 21:14:00 hi all Jul 16 21:18:14 I just need to fetch all sources needed and then do offline build how can I do so? Jul 16 21:21:50 khem: bitbake -c fetchall (i never use it much myself) Jul 16 21:21:56 khem: and hi! :-) Jul 16 21:22:15 likewise: thx Jul 16 21:22:16 khem: bitbake -c fetchall console-image is more appropriate btw Jul 16 21:22:27 likewise thanks Jul 16 21:22:42 khem: welcome, tnx for the cross toolchain rework! Jul 16 21:22:49 this is useful as sometimes I am behind strong firewalls Jul 16 21:23:09 khem: hmm, no way out or in? Jul 16 21:23:13 likewise: thx, I would appreciate if you could test some combos Jul 16 21:23:18 I know you have ppc Jul 16 21:23:37 yup, will test the three C libs for my powerpc targets Jul 16 21:23:43 likewise no in/out at all Jul 16 21:23:49 and avr32/uclibc Jul 16 21:23:57 likewise: thank you for your time Jul 16 21:24:26 khem: I will try and apply it to poky and run it through a few hoops, I just can't promise when Jul 16 21:24:27 likewise Jul 16 21:24:43 RP: Cool Jul 16 21:25:04 I am also working on getting locales build in cross env Jul 16 21:25:18 at least for eglibc Jul 16 21:25:30 we can get away from using qemu etc. for it Jul 16 21:25:53 That would be great Jul 16 21:26:18 khem: oh... that would be quite nice Jul 16 21:26:23 I have it working in my own toolchain scripts Jul 16 21:26:58 I need to translate into OE recipes someday Jul 16 21:27:21 * flo_lap thinks khem is a toolchain guru Jul 16 21:28:03 btw nptl will be in next uclibc release but I am happy that new build sequence takes care of that Jul 16 21:28:43 khem *g* Jul 16 21:28:52 next uclibc release next year Jul 16 21:29:11 woglinde_: We are working on cramming one soon Jul 16 21:29:26 khem *g* but what I read in the ml Jul 16 21:29:27 we have to get nptl branch merged into trink Jul 16 21:29:42 seems similar to git switch in oe Jul 16 21:29:55 lol Jul 16 21:30:32 btw. there was a patch on arm-linux which would fix the kernel build with gcc 4.3 Jul 16 21:30:39 I will try it tonight Jul 16 21:30:51 hm oh kernel build was broken? Jul 16 21:30:55 never noticed it Jul 16 21:31:04 no it will build but it wont boot :) Jul 16 21:31:24 hm but wasnt the fix koehn checked in? Jul 16 21:31:27 03mickeyl 07org.oe.dev * rc886586b... 10/ (1 packages/images/fso-image.bb): fso-image: remove mplayer, add mtd-utils Jul 16 21:31:28 GCC is becoming smart with every release Jul 16 21:31:32 03mickeyl 07org.oe.dev * r4c6f4640... 10/ (1 packages/images/fso-image-console.bb): add fso-image-console.bb Jul 16 21:32:02 woglinde_: might be I might have not noticed Jul 16 21:32:12 I usually use upstream kernels Jul 16 21:32:38 hm Jul 16 21:32:44 ah maybee this patch Jul 16 21:33:10 false optimazation Jul 16 21:33:38 yes infact its not false :) Jul 16 21:34:51 woglinde_: is this patch already in linux-rp-2.6.xx or somesuch Jul 16 21:35:02 dont know Jul 16 21:35:05 ask rp for p Jul 16 21:35:09 aeh linux-rp Jul 16 21:35:33 woglinde_: btw. I took care of one of your wanted patched into uclibc-nptl branch it should trickle upstream with merge Jul 16 21:35:54 khem the header-generation? Jul 16 21:36:05 hello, what browsers in openembedded repository that has the following capacities: greasemonkey,add block and media player plug-in(so i can see youtube video with greasemonkey and my media player)....firefox3?(does it have some part desactivated?) minimo? another based on webkit? Jul 16 21:36:33 gnutoo quick look at the packages dir Jul 16 21:36:40 firefox3 is in and minimo too Jul 16 21:36:51 but dont know in which shape the minimo package is Jul 16 21:37:54 woglinde_, so there is no desactivated functions in firefox3 that prevent the features i want from beeing used? i only see patches in the bb file... Jul 16 21:38:26 woglinde_, ah...mozconfig...i should look at this file... Jul 16 21:40:03 woglinde_: this thread here discussed it http://article.gmane.org/gmane.linux.ports.arm.general/10760 Jul 16 21:40:18 doesn't seem to have much desactivated...thanks a lot...i hope that firefox won't be too slow on the openmoko...moreover i should follow this guide: http://wiki.eeeuser.com/howto:shrinkfirefox Jul 16 21:41:32 * khem goes behind the firewalls Jul 16 21:42:56 bye khem Jul 16 21:43:53 khem yes Jul 16 21:44:03 this I read too Jul 16 22:01:46 03mickeyl 07org.oe.dev * r7926f38b... 10/ (4 files in 2 dirs): fso-image: add iptables Jul 16 22:19:54 03mickeyl 07org.oe.dev * r7c35ba48... 10/ (1 conf/distro/include/sane-srcrevs.inc): sane-srcrevs.inc: add dates for linux-openmoko-devel and u-boot-openmoko-devel Jul 16 23:24:27 anyone building a powerpc arch machine for KERNEL_IMAGETYPE=uImage? I find that kernel.bbclass makes in invalid test for compressed kernel for this architecture when running mkimage Jul 16 23:28:13 In troubleshooting this issue I find that my arm based machine builds a uImage without the do_deploy task defined in kernel.bbclass - that I don't quite understand Jul 17 01:51:13 how do i set the size of the jffs2 image that is going to be created? Jul 17 02:27:15 tried both ROOT_FLASH_SIZE and IMAGE_ROOTFS_SIZE_jffs2, doesn't seem to work. **** ENDING LOGGING AT Thu Jul 17 02:59:57 2008