**** BEGIN LOGGING AT Fri Feb 10 10:59:56 2006 Feb 10 11:36:40 gimp package lack icon Feb 10 11:40:43 I'm trying to figure out how to make a .bb file for my own application, but guide is somewhat useless... what about the file structure... where do I need to put the .bb file, what about my source files? Feb 10 11:42:30 ~lart libgdbm for annoying my in do_stage() Feb 10 11:42:31 * ibot breaks out the Hoover and sucks up libgdbm for annoying my in do_stage() Feb 10 11:42:57 just put my .bb file in org.openembedded.dev/packages/appname, and source in my sources dir? Feb 10 11:43:15 then try bitbake appname ? Feb 10 11:49:36 yes Feb 10 11:49:38 ;) Feb 10 11:54:33 hmmm, I'll give it a go then, if that was a yes to me... hehe... Feb 10 11:54:44 it was Feb 10 12:00:34 is 2.6 usable in ipaq3900? Feb 10 12:03:22 03hrw 07org.oe.oz354fam083 * rbaaaa5a1... 10/conf/machine/spitz.conf: Feb 10 12:03:22 spitz.conf: this machine does not use flash for things other then pivotboot Feb 10 12:03:22 so we do not need to generate JFFS2 images for it. Feb 10 12:03:22 - pivotboot change is coming. Feb 10 12:03:26 03hrw 07org.oe.oz354fam083 * rd5ea693c... 10/packages/meta/pivotboot-image.bb: Feb 10 12:03:26 pivotboot-image: on spitz build also jffs2 Feb 10 12:03:26 - by default spitz use tar.gz only Feb 10 12:03:31 03hrw 07org.oe.oz354fam083 * rce138cc8... 10/.mt-attrs: .mt-attr: drop packages/sharp-binary-only/sharp-sdmmc-support-2.4.18-rmk7-pxa3-embedix/sdmgr Feb 10 12:05:39 ~lart kopete Feb 10 12:05:39 * ibot pries kopete's back open with a screwdriver and flashes a new bootldr to kopete Feb 10 12:09:50 Is there some documentation about how to use the scirtps in the .bb files? For instance how do I make a script for a simple library? oe_libinstall "something" I'd like to read up on this... Feb 10 12:11:58 and what does the do_stage() script do? Feb 10 12:19:37 hi Bernardo Feb 10 12:22:59 am I asking stupid questions? Are there any docs on oe_libinstall ?! Feb 10 12:23:31 mndctrl: the best doc is the 3000 examples in OE Feb 10 12:24:58 * NAiL wishes someone with OE knowledge wrote some docu ;) Feb 10 12:25:27 hehe, looking at em ;) but I'm not that familiar with c lib compiling, so I don't know what oe_libinstall -a means... or all the other options... and how and where do I spesify the soname etc... Feb 10 12:37:15 hmm.. where should I tell bitbake that my app exists? I just get "ERROR: No files" when trying bitbake appname Feb 10 12:39:28 do I need to update the db or something? In oreder to get bitbake to find my app? Feb 10 12:43:25 mndctrl: try cleaning up the cache (usually inn tmp/cache) Feb 10 12:43:49 the best way even is "touch conf/local.conf" Feb 10 12:45:35 hi zecke Feb 10 12:50:26 damn tailor is upset :} Feb 10 12:52:09 http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=66c4a96409f67476191a2635d4577125b7dafa40 Feb 10 12:52:39 ~lart justin ;) Feb 10 12:52:39 * ibot executes killall -KILL justin ;) Feb 10 12:53:36 Bernardo: just remove all the .bb stuff in there? Feb 10 12:54:26 I tried the bitbake -b option, but I get: The included, default BB base.bbclass does not define a useful default task. Feb 10 12:55:37 But my .bb file is pretty much like all the rest of the libs... Feb 10 12:55:55 hmm.. not easy wthout real docs :/ Feb 10 13:12:47 lunch Feb 10 13:46:13 xfce is broken ;( Feb 10 13:46:24 xfce4-panel segfaults Feb 10 14:07:39 I figured it out ;) nevermind all my stupid questions... But I'd still appreciate better documentation for adding packages (creating .bb files, oe_libinstall etc. syntax etc. etc. ) Feb 10 14:12:40 mndctrl: create one, after you found out how to use it Feb 10 14:22:44 zecke: I'm considering it ;) Feb 10 15:02:52 someone killed ewi ;( Feb 10 15:25:19 03hrw 07org.oe.oz354fam083 * rbaecf137... 10/packages/vim/ (vim-6.4/configure.patch vim-tiny_6.4.bb vim.inc vim_6.4.bb): vim: added 6.4 from .dev (has better package split) Feb 10 15:26:43 gints|wrk: hey Feb 10 15:28:37 hi zecke Feb 10 15:46:32 hi Feb 10 15:53:40 hi Feb 10 15:55:20 hrw|work: for e packages should I prepare a patch for stable? I could easily prepare a drop/add patch Feb 10 15:55:51 hr or, if I have to, I can make it do renames and such.... Feb 10 15:55:51 hr or I can just commit them myself if that's ok with you and mickeyl Feb 10 15:56:10 hrw|work: I mean you...don't know why that other name is showing up... Feb 10 15:56:25 JustinP: commit - add note that it was discussed Feb 10 15:57:18 hrw|work: ok Feb 10 15:57:31 hrw|work: should I worry about renames or can I just drop/add Feb 10 15:57:59 I can understand if the rename sare wanted...it's just a lot of files to do that for (I'll use a script, fo course...) Feb 10 15:58:10 best way would be renames Feb 10 15:58:25 ok Feb 10 15:58:42 I'll try to do it in one commit instead of the many many I've put in dev Feb 10 15:58:50 or do we want *all* of the revisions? Feb 10 15:59:11 (I'll also wait until I've confirmed that the image is working for me in .dev) Feb 10 15:59:57 best way would be: patch .oz354fam083; make build from it;test; if work then commit Feb 10 16:00:25 yes, of course Feb 10 16:00:34 ok Feb 10 16:00:46 I'll do it in one commit then, much easier Feb 10 16:01:18 currently e-image is broken (== cannot be built) in .oz354fam083 anyway Feb 10 16:02:41 I know Feb 10 16:02:47 they removed the older packages :-( Feb 10 16:02:51 it *was* working... Feb 10 16:03:08 I know Feb 10 16:03:08 newer stuff seems to be working for me, though Feb 10 16:03:25 just trying to add a little more polish (e17-gpe-menu-convert) Feb 10 16:34:27 mickeyl: we need 480x640 pics for tosa opie-image: firstuse and background - now qvga one is used so it looks shitty Feb 10 16:45:57 cu Feb 10 17:18:26 * Philippe is back (gone 42:23:18) Feb 10 18:41:34 RP: g_ether isn't playing well when connected to a Mac either Feb 10 18:42:05 RP: it connects and I can ping and transfer, but copying files (over scp) to my Z from the mac is going at 1.5K/s or less and frequently stalling Feb 10 18:51:44 JustinP: hey :) how did you create this gtk-doc.m4 commit? Feb 10 18:52:20 JustinP: not that is your fault but you fucked up svn and hg with it ;) (at least tailor is upset) Feb 10 18:54:02 heh Feb 10 18:54:12 "create"? Feb 10 18:54:19 I just did normal monotone things Feb 10 18:54:41 renamed 2 of the gtk-doc files (moved, actually, to a new dir, "files") Feb 10 18:54:45 then deleted the rest Feb 10 18:54:57 what's breaking? Feb 10 18:55:29 JustinP: you renamed one file to a file that already exists :} Feb 10 18:55:39 no i didn't.... Feb 10 18:55:41 JustinP: so svn mv and hg mv barf out Feb 10 18:55:56 the files/ dir didn't exist previously... Feb 10 18:56:02 hmm Feb 10 18:56:12 http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=66c4a96409f67476191a2635d4577125b7dafa40 Feb 10 18:56:15 * JustinP looks at the rev Feb 10 18:56:20 at least something there upsets svn Feb 10 18:56:28 maybe it needs to commit the 'removal' first Feb 10 18:56:38 and you can't do that in one revision (which you can't) Feb 10 18:56:51 ah, I see Feb 10 18:56:54 so it is not your fault at all! Feb 10 18:57:01 I missed that files/gtk-doc.make existed Feb 10 18:57:16 oops, I shoul dhave left that alone and deleted the orbit2 one Feb 10 18:57:24 would have been less confusing Feb 10 18:57:34 but hey, I helped you find a bug in tailor ;-) Feb 10 18:57:45 any idea if it's fixable? Feb 10 19:06:39 RP: it also creates a new interface every time I plug it in....up to en6 now.... Feb 10 19:20:07 03mreimer 07org.oe.dev * radfa830f... 10/packages/h2200-bootloader/h2200-bootloader.bb: Feb 10 19:20:07 h2200_bootloader: make it a bit quieter, and save the bootloader to the more Feb 10 19:20:07 appropriate filename h2200_bootloader.bin. Feb 10 19:29:29 hi all Feb 10 19:34:09 'lo Feb 10 19:44:45 ~lart bb Feb 10 19:44:45 * ibot duct-tapes bb to the floor and drools on him Feb 10 19:52:35 03mreimer 07org.oe.dev * r8939ff9c... 10/conf/machine/h2200.conf: Feb 10 19:52:35 h2200.conf: Fix EXTRA_IMAGECMD_jffs2 to generate a working jffs2 image on Feb 10 19:52:35 h2200. Also remove unneeded module_autoload directives. Feb 10 19:52:40 03mreimer 07org.oe.dev * r0b3ce0b4... 10/conf/machine/include/handheld-common.conf: Feb 10 19:52:40 handheld-common.conf: Add hidp and irtty-sir to HANDHELD_MODULES, so these Feb 10 19:52:41 modules get included in the rootfs. Feb 10 19:53:25 mreimer: I'm not sure that we want irtty-sir in HANDHELD_MODULES. Feb 10 19:53:45 Do people really use that in any serious way? Feb 10 19:54:10 pb_: teach me; I don't know much about ir, just that I had to load it to get irattach to work Feb 10 19:54:28 mreimer: I would have thought you should be using pxa_ir instead Feb 10 19:54:49 (which, iirc, doesn't require irattach) Feb 10 19:54:58 pb_: ok, I'll back that out and investigate. thanks for piping in Feb 10 19:55:32 irtty-sir is kind of the irda driver of last resort: it works on pretty much any generic serial port, but you only get SIR speeds (i.e. 9600bps) Feb 10 19:55:40 pb_: what about hidp? is that ok? Feb 10 19:55:45 yeah, hidp is good Feb 10 20:14:45 pb_: is CONFIG_PXA_FICP what you were referring to? Feb 10 20:15:55 I'm not sure what it's called in 2.6. FWIW, the corresponding 2.4 option is: Feb 10 20:15:57 dep_tristate 'Intel PXA2xx Internal IR' CONFIG_PXA_FIR $CONFIG_IRDA $CONFIG_EXPERIMENTAL Feb 10 20:17:25 ibot, seen lrg? Feb 10 20:17:41 lrg was last seen on IRC in channel #oe, 35d 3h 35m 38s ago, saying: 'Luke-Jr: alsa now gives the flexability for additional controls that don't easily fit into the oss model'. Feb 10 20:17:50 doh Feb 10 20:53:27 where does auto-munging of package names happen? Feb 10 20:53:36 e.g. embryo => libembryo Feb 10 21:03:55 not sure Feb 10 21:06:08 does anyone here have any experience with using qemu to run a native ARM compiler on an x86 host to build things for the Zaurus? Feb 10 21:09:58 ....can you do that? Feb 10 21:10:09 I *think* so Feb 10 21:10:19 I'm looking at something called scratchbox Feb 10 21:12:02 03mreimer 07org.oe.dev * rb6ccd39b... 10/conf/machine/include/handheld-common.conf: handheld-common.conf: Remove irtty-sir from HANDHELD_MODULES. Feb 10 21:15:17 evening Feb 10 21:15:33 hello Feb 10 21:16:25 hi reenoo Feb 10 21:16:27 hi johnX Feb 10 21:16:29 hey pb_ Feb 10 21:18:26 pb_: seems like someone owes you a beer: http://www.fsfe.org/en/fellows/robertschuster/weblog/hacking_for_beer http://classpath.wildebeest.org/dairy-man-di/index.php?p=30 Feb 10 21:20:16 heh Feb 10 21:37:23 Greetings ! Feb 10 21:37:30 yo Feb 10 21:38:06 i am having a strange problem with bitbake.... for somereason it seems to "ignore" the BBFILES settings in local.conf Feb 10 21:38:34 i need to export the BBFILES dir in order to make it work.... Feb 10 21:38:55 any ideas ? Feb 10 21:39:39 it's not installing....WTF? what happened to my build? Feb 10 21:40:22 Ifaistos: are you exporting BBPATH? Feb 10 21:41:09 yes. it used to work.... things started to fall apart once i moved my working dir Feb 10 21:41:34 mmhm Feb 10 21:41:47 Ifaistos: sounds like your local.conf isnt benig parsed at all, which indicates that your BBPATH is incorrect. Feb 10 21:42:07 i am preperaing the dirs for an svn import and moved everything to owmnr-svn/trunk/ Feb 10 21:42:47 Ifaistos: did you empty your cache? Feb 10 21:42:54 Ifaistos: did you set your env vars right? Feb 10 21:43:24 Ifaistos: echo $BBPATH Feb 10 21:44:33 i think BBPATH is what is causing the trouble.... Feb 10 21:44:37 stelios@Athena:~/owmnr-svn/trunk$ echo $BBPATH Feb 10 21:44:37 /home/stelios/owmnr-svn/trunk:/home/stelios/owmnr-svn/trunk/org.openembedded.dev:/home/stelios/owmnr-svn/trunk/bitbake Feb 10 21:45:07 the think is that i am building on trunk/x86/build Feb 10 21:45:09 does /home/stelios/owmnr-svn/trunk/conf/local.conf exist? Feb 10 21:45:17 which is not in BBPATH Feb 10 21:45:22 you probably wanted trunk/build in bbpath Feb 10 21:45:23 right Feb 10 21:46:55 shadows, JustinP and kergoth many thanks ! Feb 10 21:47:09 sure. keep on truckin' Feb 10 21:48:12 i am working on a distro for a wireless router, to be used in our comunity base wifi network Feb 10 21:48:17 ~lart evas-x11 Feb 10 21:48:17 * ibot overclocks evas-x11 until evas-x11 burns out Feb 10 21:48:36 http://wind.awmn.net Feb 10 21:49:43 you better try this url or they will look Greek to you :) http://wind.awmn.net/?page=nodes Feb 10 21:50:31 Ifaistos: nice map :-) Feb 10 21:58:23 hmm what do you all think of the Palm Treo as a hardware target Feb 10 22:00:41 does it run Linux? Feb 10 22:03:25 yeah. it apparently can. Feb 10 22:03:30 http://www.grack.com/programming/misc/TreoLinux.html Feb 10 22:03:32 and GPE Feb 10 22:03:36 ;) Feb 10 22:03:38 Wow- somebody hacked the Treo? Feb 10 22:04:04 yeah. it was news to me Feb 10 22:04:18 l33t Feb 10 22:04:21 i'm asking this guy on their IRC resource "luke" and he is saying that they use OE Feb 10 22:04:24 it's not integrated yet Feb 10 22:04:57 I'm sure patches would be welcome Feb 10 22:16:34 * Philippe is away: visual contact - Helsinki productivity killer Feb 10 23:20:29 pb_: ping. question about cairo bb's Feb 10 23:20:37 they need some love. Feb 10 23:25:46 there's only libsvg_cvs, no libsvg_0.1.4 Feb 11 00:33:01 hi people Feb 11 00:34:42 somebody know what happened to openzaurus.org domain? a nslookup gives not existant entries Feb 11 00:35:14 the bill didnt get paid. will take care of it shortly Feb 11 00:36:38 oh, i see Feb 11 00:36:57 is the project hosted anywhere else? Feb 11 00:37:37 kergoth: could you *please* change the #oz topic to reflect the websites for bugtracker, openzaurus, ? Feb 11 00:37:40 on #openzaurus shadow answered that Feb 11 00:38:05 is't hosted on openzaurus.sf.net Feb 11 00:41:24 is anyone around who can try to build evas-x11 for me? Feb 11 00:42:17 JustinP: I can, later. I'm building in .dev from scratch, on glibc now Feb 11 00:44:42 mreimer: I'd appreciate it Feb 11 00:44:51 mreimer: let me know if it packages the main package Feb 11 00:45:05 JustinP: ok. it will be a while Feb 11 00:45:11 mreimer: and if it build evas-x11 packages or libevas packages Feb 11 00:46:53 on my way out Feb 11 01:03:11 * JustinP is back Feb 11 01:04:48 mreimer: I just started a new build from the "glibc done" point Feb 11 01:05:04 mine is currently packing glibc Feb 11 01:05:31 I'm having a very strange problem Feb 11 01:05:32 erg. libexpat versus libxml Feb 11 01:05:38 what should be used? Feb 11 01:05:47 expat is smaller Feb 11 01:05:58 evas-x11 worked fine yesterday but today the splitting isn't working right.... Feb 11 01:06:14 NOTE: package gcc-cross-3.4.4: started Feb 11 01:06:17 gonna be a while Feb 11 01:06:41 I have that done too :-) Feb 11 01:06:50 I keep a copy of tmp from after gcc-cross Feb 11 01:07:20 o-kay, need to add in expat-2.0.0 then Feb 11 01:11:25 mreimer: still going to take forever for me...have to compile diet-x11 and all Feb 11 01:11:35 just finished jpeg, though ^_^ Feb 11 01:11:49 maybe by tomorrow :-) Feb 11 01:11:56 oooh, already starting eet. You can get efl started ptretty quick Feb 11 01:12:11 still on gcc-cross :-( Feb 11 01:13:00 yeah, glibc and gcc-cross are the worst 2 Feb 11 01:13:18 *cough* binutils *cough* Feb 11 01:13:27 is that one long too? Feb 11 01:13:38 * JustinP hasn't noticed Feb 11 01:13:58 if you build all possible targets, i think it may be the longest GNU package to build Feb 11 01:14:19 t2 uses it as a reference case Feb 11 01:14:27 ok Feb 11 01:14:37 well, I'm not building for all targets....why would I want that? ;-) Feb 11 01:14:51 :) Feb 11 01:19:10 not to figure out this other problems.... Feb 11 01:19:13 -s Feb 11 01:19:44 module-init-tools-cross Feb 11 01:20:06 ? Feb 11 01:20:11 ncurses Feb 11 01:21:13 | /usr/bin/install: cannot create regular file `/usr/share/man/man1/xmlwf.1': Permission denied Feb 11 01:21:16 derh? Feb 11 01:21:23 I started this build for handhelds-pxa-2.6 Feb 11 01:21:45 mreimer: which machine is that? Feb 11 01:21:51 shadows: did it kill the build? Feb 11 01:22:06 JustinP: my build host? sempron 2400 (1.6GHz) Feb 11 01:22:10 i don't know what i'm doing, really. i was trying to add a bb for expat 2.0.0 Feb 11 01:22:20 using the previous expat bb version as a template Feb 11 01:22:44 doesn't autotools_do_configure handle the man page configuration var? Feb 11 01:23:36 shadows: it should.....but there are times when the makefiles aren't set up right Feb 11 01:23:51 mreimer: I meant what machine is the kernel for Feb 11 01:24:00 JustinP: h2200 Feb 11 01:24:27 JustinP: i see no mention of 'mandir' in the output of the log file Feb 11 01:24:43 err Feb 11 01:24:45 no there it is Feb 11 01:24:55 --mandir=/usr/share/man Feb 11 01:25:13 so erm, what am i supposed to be making this do Feb 11 01:25:47 with Portage we had a sandbox, and wrapper functions and complexity Feb 11 01:26:05 bitbake/OE has complexity too Feb 11 01:26:18 you can try looking at the autotoolsbbclass and such Feb 11 01:26:24 actually, that mandir is probably right Feb 11 01:26:31 yeah Feb 11 01:26:32 the DESTDIR is altered on install I think Feb 11 01:26:40 oh okay, so check DESTDIR fooby Feb 11 01:26:47 I think Feb 11 01:26:57 I'm not a master of autoconf and friends Feb 11 01:27:10 only a few gleaned half-truths.... Feb 11 01:27:43 right now I'm trying to debug why my efl includes are installed in /home/papercrane/oe... within the image dir... Feb 11 01:28:57 what the heck is INSTALL_DATA Feb 11 01:29:06 don't know Feb 11 01:29:07 * shadows pokes expat-2.0.0 source Feb 11 01:29:22 configure:19359:test -z "$INSTALL_DATA" && INSTALL_DATA='${INSTALL} -m 644' Feb 11 01:29:40 what the hell? how to procede? Feb 11 01:30:56 going to try autotoolsbbclass Feb 11 01:31:24 s.bb Feb 11 01:31:28 no, that's already included Feb 11 01:31:28 I forgot a . Feb 11 01:31:39 yeah, but you might look into it for clues Feb 11 01:31:44 inherit autotools lib_package Feb 11 01:31:46 hmn Feb 11 01:31:57 diet-x11, woo Feb 11 01:32:17 shadows: I mostly resorted to sed hackery to fix my autofoo problems Feb 11 01:32:21 oh Feb 11 01:32:30 should /usr/bin/install be invoked? Feb 11 01:32:36 that's the point i'm confused about Feb 11 01:32:42 how are these files supposed to be installed Feb 11 01:32:53 ? Feb 11 01:33:10 install is just a file copier which can set attrs Feb 11 01:33:24 install is used to install the files to the "image" dir Feb 11 01:33:24 yes, but should /usr/bin/install from the *host machine* be invoked Feb 11 01:33:35 ::shrug:: I don't see why not Feb 11 01:33:59 I don't believe a version is compiled.... Feb 11 01:34:03 of course I'm not sure Feb 11 01:42:51 on xext now Feb 11 01:42:55 won't be too long Feb 11 01:43:01 * JustinP needs a faster machine for building Feb 11 01:43:39 it seems that something like libpng uses "make -e MAKEFLAGS= ... install" Feb 11 01:45:26 there goes evas-x11 :-) Feb 11 01:45:41 damn,....I know what's breaking the includes....it's my sed hackery :-| Feb 11 01:50:21 patch good Feb 11 01:50:22 sed bad Feb 11 01:51:44 well, evas-x11 still packaged wrong for me, so it wasn't my cache and all Feb 11 01:51:48 :-( Feb 11 01:52:04 * JustinP moves back his other tmp Feb 11 01:53:10 i think i found how to make expat-2.0.0 behave Feb 11 01:53:18 needed to add a line to define man1dir Feb 11 01:55:44 NOTE: the following files were installed but not shipped in any package: Feb 11 01:55:44 NOTE: xmlwf.1 Feb 11 01:55:49 what the heck does that mean? Feb 11 01:56:14 it means that: Feb 11 01:56:22 1) it was installed in the / dir, which is wrong Feb 11 01:56:27 ah Feb 11 01:56:28 2) that is was not packages Feb 11 01:56:33 packaged Feb 11 01:56:54 the path displayed in that message is the path from the "image" or "sandbox" install dir Feb 11 01:56:54 so like, it was installed in the root of the build dir? errr.. Feb 11 01:57:04 oh okay, groking logs Feb 11 01:57:05 meaning, it's the full path after a hypothetical install Feb 11 01:57:59 export mandir="/usr/share/man" Feb 11 01:58:01 from the log Feb 11 01:58:02 hm Feb 11 01:58:29 man1dir="/home/jnc/opensource/openzaurus/build/oetmp/work/armv5te-linux/expat-2.0.0-r0/image${man1dir}" Feb 11 01:58:32 from inside do_install Feb 11 01:58:35 that looks incorrect Feb 11 01:58:58 my guess is i should change this to closer match zlib, "man3dir=${D}${mandir}/man3" Feb 11 01:59:28 okay, updated, going to try again Feb 11 02:03:01 works! Feb 11 02:03:04 * shadows :) Feb 11 02:03:49 shadows: want to take a look at my problem? Feb 11 02:04:10 sounds like you have more autofoo karma than I do Feb 11 02:04:11 ;-) Feb 11 02:04:36 hehe Feb 11 02:04:38 sure? Feb 11 02:04:48 well, I have 2 issues right now... Feb 11 02:04:58 can you try building evas-x11? Feb 11 02:05:35 yeah, it's going to be with gcc-4.0.2 cross though Feb 11 02:05:42 so that is kind of incompatible at the moment Feb 11 02:05:55 hmmmm...true....but it might work... Feb 11 02:06:03 urgh, probably not, though Feb 11 02:06:14 let me resolve a thing or two with expat, which is the current hangup, then i'll go bugfix towards evas-x11 Feb 11 02:06:19 ok Feb 11 02:06:23 thx Feb 11 02:06:26 should i be setting PR ? Feb 11 02:06:33 i see some builds do, some do not Feb 11 02:06:37 yes Feb 11 02:06:40 always set a PR Feb 11 02:06:42 okay Feb 11 02:06:43 PR = "r0" Feb 11 02:06:54 then up it whenever you make a change (and commit it) Feb 11 02:06:57 if i see a build that does not set a PR, should i correct it? Feb 11 02:06:58 so that upgrades will work Feb 11 02:07:05 i don't have commit access Feb 11 02:07:08 all build *should* have one..... Feb 11 02:07:24 (unless i am unaware of something) Feb 11 02:07:25 if they don't (and no incude or inherit sets one) then file a bug...I think Feb 11 02:07:29 okay Feb 11 02:07:32 sorry, wasn't aware of that Feb 11 02:07:41 it's better that way, for now Feb 11 02:07:50 perhaps if you do more gcc4 stuff you'll get commit :-) Feb 11 02:07:54 i'm prone to break things in the name of newer versions heh heh Feb 11 02:07:57 yeah Feb 11 02:08:03 yeah, that's bad :-| Feb 11 02:08:10 i am supposed to write a WPA HOWTO for oz.org Feb 11 02:08:18 I got it because I volunteered to maintain e17 Feb 11 02:08:24 :-) Feb 11 02:08:27 hopefully i can keep OZ running on my Z long enough to get through a HOWTO writeup Feb 11 02:08:39 currently waiting for a serial cable i bought Feb 11 02:08:51 $50usd for the sharp Z SIO cable, not bad eh? Feb 11 02:08:56 not bad, no Feb 11 02:09:17 they're going on eBay for 25.00 + 24.99 s/h Feb 11 02:09:21 SIO? Feb 11 02:09:32 SharpIO, a type of legacy port on sharp hardware Feb 11 02:09:45 includes historically the serial, JTAG, power, and USB Feb 11 02:09:48 ah Feb 11 02:10:14 i got one of those with my collie, didnt know they werent always included Feb 11 02:10:35 i think it was standard with some collie sales Feb 11 02:10:51 the cable is difficult to acquire now that the serialio.com cable is popular Feb 11 02:11:07 the serialio.com cable does not properly convert levels, resulting in one-way communication only Feb 11 02:11:29 i see Feb 11 02:11:43 it's broken for the newer models Feb 11 02:11:46 works for older ones Feb 11 02:11:56 * JustinP was a sucker and bought one Feb 11 02:12:14 JustinP: you should be able to "fix" the level conversion Feb 11 02:12:17 it's actually out of spec Feb 11 02:12:29 that's why the serialio.com cable is not functional Feb 11 02:13:07 you mean the zaurii are out of spec? Feb 11 02:13:29 I'm not surprised, it's not meant to be a functional port (user-side, that is) Feb 11 02:13:30 i think that may also be accurate Feb 11 02:13:42 not in the newer ones Feb 11 02:13:46 at least Feb 11 02:13:51 it's a sort of TTL port Feb 11 02:14:43 i've got a URL which has this documented for sharp calculators, which may be relevent: http://my.ebay.com/ws/eBayISAPI.dll?MyeBay Feb 11 02:15:22 i think ultimately it would be in our best interests to discover how to craft and use a jtag cable for sharp hardware Feb 11 02:15:34 it doesn't seem so difficult, if we can find a source for the connectors Feb 11 02:15:55 the rest is just oscilloscope kung-fu'ery Feb 11 02:16:20 I'm pretty sure people have done it... Feb 11 02:16:56 none i have heard of, only a few for SL-5500 and those were people with 20+ years of business contact with Sharp Feb 11 02:17:49 ah Feb 11 02:17:54 i'm tempted to place an order for 1k pcs of that connector type and ebay it to hell Feb 11 02:18:01 charge 10 bucks a connector ;) Feb 11 02:18:59 shadows, some devices get convoluted trying to find an exposed trace to connect a jtag connector Feb 11 02:19:13 hmm :/ Feb 11 02:19:15 part of thier "physical security" Feb 11 02:19:48 cellphones are a prime example Feb 11 02:20:12 you used to have dealer "programming" connectors exposed behind the batteries Feb 11 02:20:45 most cell companies have stopped doing that since people were using them to clone abd reprogram phones Feb 11 02:20:50 reprogram* Feb 11 02:21:00 ah Feb 11 02:25:18 http://bugs.openembedded.org/show_bug.cgi?id=672 Feb 11 02:25:23 expat update to 2.0.0 Feb 11 02:25:30 who wants to commit, you know you do! Feb 11 02:25:49 as miyavix_visavis would say, the answer is in you Feb 11 02:28:22 is bugs finally back up? Feb 11 02:28:43 nope Feb 11 02:32:45 add a line to your hosts file, 192.216.230.225 bugs.openembedded.org Feb 11 02:33:03 it will keep your sanity engaged. Feb 11 02:42:02 * france is back (gone 18:54:44) Feb 11 02:43:38 * JustinP hacks package.bbclass Feb 11 02:45:08 lol Feb 11 02:45:45 trying to figure out where it's killing my damn package Feb 11 02:46:00 the -config thing still ? Feb 11 02:47:37 no Feb 11 02:47:39 fixed that Feb 11 02:47:55 bitbake treats FILES_${SRCNAME}-dev and FILES_${PN}-dev differently Feb 11 02:48:03 even though they may to the same thing.... Feb 11 02:48:16 may? Feb 11 02:48:22 map Feb 11 02:48:37 I've got a very random problem now.... Feb 11 02:48:42 evas-x11 isn't getting renamed to libevas Feb 11 02:48:52 and the main package isn't being built as it's "empty" Feb 11 02:49:12 grrrrr! Feb 11 02:49:22 the package evas-x11 is being populated twice Feb 11 02:50:00 ... Feb 11 02:50:19 how fun Feb 11 02:50:19 NOTE: packages = evas-x11 evas-x11-doc evas-x11-dev evas-x11-locale evas-x11 evas-x11-themes evas-x11-dev evas-x11-examples Feb 11 02:50:22 NOTE: pkg = evas-x11 Feb 11 02:50:25 NOTE: pkg = evas-x11-doc Feb 11 02:50:27 NOTE: pkg = evas-x11-dev Feb 11 02:50:30 NOTE: pkg = evas-x11-locale Feb 11 02:50:32 NOTE: pkg = evas-x11 Feb 11 02:50:35 NOTE: pkg = evas-x11-themes Feb 11 02:50:37 NOTE: pkg = evas-x11-dev Feb 11 02:50:40 NOTE: pkg = evas-x11-examples Feb 11 02:50:42 aha Feb 11 02:50:45 it's my PACKAGES += Feb 11 02:50:47 still don't know why evas-x11 doesn't work and embryo does... Feb 11 02:50:50 * JustinP apologizes for spamming Feb 11 02:50:52 lol Feb 11 02:53:27 is it possible to make packaes.split() also make the entries unique? Feb 11 02:53:28 hmm Feb 11 02:53:45 cmon edje ... go transparent Feb 11 02:58:08 ibot: botmail for mickeyl, how can I make an array in python have only unique values? Feb 11 03:01:21 do a search for duplicates i'd think Feb 11 03:06:47 oh hell yes Feb 11 03:07:03 of course I can't *commit* this change.....at least not without discussion Feb 11 03:07:20 about damn time I figured out this problem Feb 11 03:07:58 ok, bbiab Feb 11 03:13:40 hooray for Justin Feb 11 03:45:04 ~lart package.bbclass Feb 11 03:45:05 * ibot blasts package.bbclass to oblivion with a kamehameha wave Feb 11 03:57:15 ibot: botmail for mickeyl, I figured it out. I have a patch for package.bbclass to use a set to make the PACKAGES unique Feb 11 04:21:54 JustinP: bug #? Feb 11 04:23:04 shadows: for? Feb 11 04:23:18 your doodle thinger Feb 11 04:23:25 doodle? Feb 11 04:23:33 patch? Feb 11 04:23:55 ummmmm..I should make one Feb 11 04:23:57 one sec Feb 11 04:29:57 it helps if you add bugs.openemebedded.org to your hosts file Feb 11 04:36:01 shadows: bugs.treke.net Feb 11 04:36:03 shadows: http://bugs.treke.net/show_bug.cgi?id=674 Feb 11 04:36:34 :) Feb 11 04:37:00 it's simple as adding bugs.openembedded.org to your hosts file Feb 11 04:37:04 makes everything work fine Feb 11 04:37:07 * shadows sighs Feb 11 04:37:14 http://bugs.openembedded.org/show_bug.cgi?id=675 Feb 11 04:37:21 updates for libsvg Feb 11 04:37:33 the patch to make gcc4 eat it wasn't so bad Feb 11 04:37:35 only two lines Feb 11 04:38:32 I don't have time to submit them right now.... Feb 11 04:38:39 JustinP: the ideal thing would be to warn and halt Feb 11 04:38:46 so that duplicate cases are fixed Feb 11 04:38:58 I suppose Feb 11 04:39:30 I can think of ok reasons to have duplicates....but I suppose it should be "fixed"... Feb 11 04:39:36 ::shrug:: Feb 11 04:39:48 as long as it doesn't just eat packages like now Feb 11 05:41:37 gah Feb 11 05:41:42 libsvg build blew up Feb 11 06:07:34 how do I set oz354fam083? Feb 11 06:10:40 it's decided when you checkout your branch Feb 11 06:11:04 i.e. oz354fam083 is one of the branches in the oe monotone scm Feb 11 06:11:28 use --branch=... i think Feb 11 06:11:38 okay, I'll try that Feb 11 06:12:07 nope Feb 11 06:12:38 I guess I'm meant to use the bleeding edge instead of the cutting edge Feb 11 06:18:13 Zero_Chaos: monotone --db=OE.db --branch=org.openembedded.oz354fam083 Feb 11 06:18:45 JustinP: can I use the same OE.db for that as well as .dev? Feb 11 06:18:54 yes Feb 11 06:18:59 it should have both Feb 11 06:19:05 really? that is wicked Feb 11 06:19:07 if you were pulling both.... Feb 11 06:19:08 thanks Feb 11 06:19:22 I've never pulled both... is there a faq somewhere? Feb 11 06:19:27 a setup guide? Feb 11 06:19:37 when you pull what do you do? Feb 11 06:20:32 monotone --db=/home/zaurus/src/build/conf/oe.db pull dominion.kabel.utwente.nl org.openembedded.dev Feb 11 06:20:39 that's the problem then Feb 11 06:20:52 you weren't pulling the other branch Feb 11 06:20:57 does the checkout fail? Feb 11 06:21:08 that command completes fine Feb 11 06:21:21 what should I do? Feb 11 06:21:40 you should then have a checkout Feb 11 06:21:54 however, you need to pull the latest for that branch Feb 11 06:22:06 monotone --db=/home/zaurus/src/build/conf/oe.db pull dominion.kabel.utwente.nl org.openembedded.oz354fam083 Feb 11 06:22:38 okay, so I can have my line, and the one you just wrote in my update script, right? Feb 11 06:22:55 yes, but don't Feb 11 06:23:00 use monotone --db=/home/zaurus/src/build/conf/oe.db pull dominion.kabel.utwente.nl org.openembedded."*" Feb 11 06:23:03 with the quotes Feb 11 06:23:08 will pull both :-) Feb 11 06:23:22 JustinP: only both, or 100 branches that I don't want? Feb 11 06:23:46 there are only 3 Feb 11 06:23:56 and .dreambox barely changes Feb 11 06:24:21 JustinP: I'm very limited on space, can I pull each individually, or is that bad? Feb 11 06:24:28 I don't want anything I don't need Feb 11 06:24:33 Zero_Chaos: use one db Feb 11 06:24:49 Zero_Chaos: if you're worried, you *can* pull them seperately... Feb 11 06:24:53 Zero_Chaos: or you can use, say: Feb 11 06:25:13 Zero_Chaos: org.openembedded."{dev,oz354fam083}" Feb 11 06:25:17 I *think* that's right Feb 11 06:25:25 I'll check Feb 11 06:26:42 JustinP: I did it with no quotes, and it seems to be working.... Feb 11 06:26:56 ok Feb 11 06:27:07 as long as it's oulling revs you should be ok Feb 11 06:27:10 * Zero_Chaos bows to the sage advice of JustinP Feb 11 06:27:24 I sure seem to be. Feb 11 06:27:51 and it will probably take a while, depending on how long you haven't been pulling it Feb 11 06:27:57 how many revs? Feb 11 06:27:59 now I can just change my conf/local.conf to use org.openembedded.oz354fam083 instead or .dev and it should work right....right? Feb 11 06:28:10 JustinP: 4321 Feb 11 06:28:19 Zero_Chaos: pretty much Feb 11 06:28:24 JustinP: Revs in 390 Feb 11 06:28:30 Zero_Chaos: you should empty your tmpdir first Feb 11 06:28:40 I imagine those revs are for the new branch Feb 11 06:28:43 Zero_Chaos: or use another one Feb 11 06:28:45 yes Feb 11 06:28:49 JustinP: yeah, I figured Feb 11 06:30:33 JustinP: thanks for your help, time for sleepy Feb 11 06:30:35 later Feb 11 06:34:37 Zero_Chaos: sleep well Feb 11 07:13:32 mreimer: i solved my icu build problem, it was a LC_ALL issue =P Feb 11 07:45:26 could i please get some people testing http://bugs.openembedded.org/show_bug.cgi?id=540 Feb 11 07:45:51 need volunteers with various targets and build platforms (amd64/ia32) Feb 11 07:54:04 not sure how many people actually have 32/64 systems Feb 11 07:55:14 either one is fine Feb 11 07:55:23 i want to verify what i'm doing is correct Feb 11 07:55:41 ... Feb 11 07:55:43 the trouble is that sqlite assumes a lot of things unless you tell it otherwise, during the compile phase Feb 11 07:55:55 says its for 32/64 systems not 32 OR 64 Feb 11 07:55:59 it makes stupid assumptions about the size of a pointer Feb 11 07:56:14 amd64 and ia32 are exclusive Feb 11 07:56:41 so your trying to test a hybrid patch on pure systems? Feb 11 07:56:47 no Feb 11 07:57:08 i need it tested on whatever platforms people have, especially if they have both platforms to test on Feb 11 07:57:27 i'm trying it with zaurus c3000 target on amd64 host Feb 11 07:57:31 more tests welcomed :) Feb 11 07:57:57 for ia32 build hosts i want to verify that the patch is functionally equivillent to no patch Feb 11 07:58:14 then i apperently dont understand what the bug is for Feb 11 07:58:21 and for amd64 build hosts i want to verify that it fixes the bug where libsqlite crashes and in a chain of unfortunate events causes tiny fonts Feb 11 07:58:22 ah Feb 11 07:59:04 I can test it later...busy right now Feb 11 07:59:16 remind me tomorrow or something... Feb 11 07:59:48 sqlite has some code in the makefile that runs using the build host's cc, and executes that code on the build host, spitting out the size of a pointer of the build host environment onto stdout, which is then directed into a config.h file and used as a value for the target source compile. totally bollocks Feb 11 07:59:54 super :) Feb 11 08:00:53 emte: what the patch i've made does is explicitly sets a makefile define per the comments of sqlite, such that it defines the type which would otherwise be set by the bollocks'd code Feb 11 08:01:06 a workaround, and should be effective if i got the type correct Feb 11 08:01:36 '-DINTPTR_TYPE int' Feb 11 08:01:40 yep Feb 11 08:01:41 your refering to that i take it Feb 11 08:02:07 yeah that wont do anything on a 32bit host Feb 11 08:02:47 32bit host, 32bit target Feb 11 08:02:56 it would have to be modified if we had a 64bit target Feb 11 08:03:02 for 64bit host i'm hoping that works okay Feb 11 08:04:13 i'm running through a gcc4 (4.0.2) compile on amd64 build host of gpe-image Feb 11 08:04:27 so far i've patched up expat and libsvg Feb 11 08:04:36 and now hopefully sqlite Feb 11 08:05:33 i am curious Feb 11 08:05:42 why dont you just test the host and adjust? Feb 11 08:06:06 emte: the test should be for the target, not the host Feb 11 08:07:09 emte: take a look at the code for sqlite if you have any further questions, i am open for suggestions Feb 11 08:07:10 but INT can vary is size from host to host even in ansii standards Feb 11 08:07:19 in* Feb 11 08:09:05 ANSI Feb 11 08:09:43 older standards define it as a unit of work dependent on CPU type Feb 11 08:09:47 yeah the extra i was for fun Feb 11 08:09:50 :) Feb 11 08:10:33 what do you think about moving oe's default compiler to something more recent Feb 11 08:10:36 like gcc 4.0.2 ? Feb 11 08:10:54 bad idea right now Feb 11 08:11:02 what depends on gcc3 ? Feb 11 08:11:06 i'd wait til 4.x was a bit more mature Feb 11 08:11:14 i think gcc4 is mature Feb 11 08:11:15 most code Feb 11 08:11:27 most code meaning what? Feb 11 08:11:29 i remember all the problems with the first few releases of 3.x Feb 11 08:11:46 legacy programming practices Feb 11 08:11:59 and non-maintianed code Feb 11 08:12:04 maintained* Feb 11 08:12:21 OE is already dependent on gcc 2.95 Feb 11 08:12:26 no Feb 11 08:12:31 which is pretty freaking ancient Feb 11 08:12:41 the sharp compiler is 2.95 Feb 11 08:12:44 yeah, for 2.4.20 oz embeddix kernels Feb 11 08:12:45 mmhm Feb 11 08:12:55 thats the only thing that uses 2.95 Feb 11 08:13:05 out of the whole OE tree, that's the only one? Feb 11 08:13:11 yup Feb 11 08:13:14 that i know of Feb 11 08:13:17 interesting Feb 11 08:13:30 familiar and OZ use 3.3.4 for the most part Feb 11 08:13:47 and i am pretty sure all the other projects do as well Feb 11 08:15:24 whee ... 1'st abstarction test passed Feb 11 08:15:36 the ra one too Feb 11 08:17:10 thanks for the feedback Feb 11 08:17:17 sleepytime for me :) Feb 11 08:18:07 yeah, i'll try to rememebr to get the patch after monotone is done Feb 11 08:33:42 JustinP: I'm afraid I have no access to macs. You could try disabling the DMA code in the driver though (its a #define near the top of the file) **** ENDING LOGGING AT Sat Feb 11 10:59:56 2006