**** BEGIN LOGGING AT Mon Jan 29 02:59:59 2007 Jan 29 04:31:52 03rwhitby 07org.oe.dev * r5e2224ca... 10/ (9 files in 5 dirs): ixp4xx-kernel: updated to the latest svn patches, and enabled the defconfig and patches series file to be overridden by local OE metadata Jan 29 04:31:58 03rwhitby 07org.oe.dev * r4babe1ca... 10/ (1 packages/images/slugos-image.bb): slugos-image: Removed Apex from the image Jan 29 04:50:51 Hmm.. any .bb hackers around? Jan 29 06:02:35 koen|gprs_: Do you know offhand if there's a way to specify a version dependency in DEPENDS? e.g. DEPENDS="slang >= 2.0.0" Jan 29 06:10:25 NAbyss: DEPENDS = "foo-56.2.2.4" Jan 29 06:10:44 koen|gprs_: Thanks Jan 29 06:11:12 DEPENDS = "foo (>= 3.443.43)" should also get parsed iirc Jan 29 06:21:31 pb: when building MESA my build dies. Sed doesn't follow the sym link you create to the configuration file. Jan 29 06:22:22 is there something wrong with my version of sed? I'm using FC6 and 4.1.6. Jan 29 06:22:42 I modify the MESA bb file to cp instead of ln and it works. Jan 29 06:31:33 RP, I mean (not pb woops) Jan 29 07:42:32 hi Jan 29 07:42:43 mornin' Jan 29 07:42:44 :> Jan 29 07:42:55 XorA|gone: I built oz 3.5.5 images on my amd64 without those patches Jan 29 07:42:55 hey hrw Jan 29 08:09:34 http://home.ctlnet.com/~minimechadon/minimechadon.htm Jan 29 08:10:11 ok. time to go to conecto Jan 29 08:10:14 3 days left Jan 29 08:10:54 too bad that they have so fscked network connection;( Jan 29 08:16:08 Hello, I just did an update, and now I get a build error on the mesa package. I was just wondering, why doesn't OE have a "stable" package that 'just builds' Jan 29 08:17:43 or is this a "manpower" problem? Jan 29 08:19:22 thegve: Was someone else earlier who also had a problem with mesa: Jan 29 08:19:22 07:53 I modify the MESA bb file to cp instead of ln and it works. Jan 29 08:19:41 ok, tnx Jan 29 08:20:18 I don't know anything more, but there's some others here who do :) Jan 29 08:21:34 ok Jan 29 08:21:40 I think it worked Jan 29 08:21:44 it is now compiling Jan 29 08:22:06 And it errored on configure before ( I think , it's early...) Jan 29 08:22:26 Yes, was something about sed being unable to follow a link Jan 29 08:22:47 to the config file Jan 29 08:26:52 morning Jan 29 08:27:52 mornin' :> Jan 29 08:32:42 Goodmorning Jan 29 08:44:19 re Jan 29 08:44:57 is anyone apart from tmbinc working the dreambox branch? Jan 29 08:45:56 hey hrw|work Jan 29 08:53:31 * XorA tries to find the magic that makes intltool-native compile Jan 29 08:54:03 morning all Jan 29 08:56:48 Hmm, yet another new build error. xserver-xorg-X11. In task do_package: "No such file or directory: ' ./usr/lib/X11/ (.........)/SPSPARC2/fonts' Jan 29 08:57:06 Did not have this error before the update this morning Jan 29 08:57:44 Don't know if someone mentioned this on this channel this morning before... I'll check the bugtracker Jan 29 09:08:02 ~lart XML::Parser... Jan 29 09:08:02 * ibot duct-tapes XML::Parser... to the floor and drools on him Jan 29 09:13:14 hrw|work: I see you are landing at Charleroi Jan 29 09:13:27 hrw|work: still debating that route from UK Jan 29 09:32:04 thegve: yeah, the xserver-xorg thing does seem broken. I'll have a look at that. Jan 29 09:34:10 morning all Jan 29 09:34:16 time to go to bed :> bye everyone Jan 29 09:34:17 hey RP pb__ Jan 29 09:34:40 hi XorA, rp Jan 29 09:35:29 I asked this earlier this morning too, but what is the reason OE does not have different branches, like debian and most distros. Like "stable" and "testing" or so. As I have the feeling there only exists a "testing" branch now (nofi, I am sure you'll all do your best to avoid errors) Jan 29 09:37:41 thegve: because OE isnt a distro Jan 29 09:37:55 thegve: oz354x is an example of a "stable" branch Jan 29 09:38:16 Ok.... Jan 29 09:38:40 But it is an os-building platform, a software package. Jan 29 09:38:54 And these often have testing and stable versions too... Jan 29 09:48:00 03pb 07org.oe.dev * r77212e45... 10/ (1 classes/package.bbclass): Jan 29 09:48:00 package.bbclass: add workaround for wrongly globbed symlinks in FILES causing python Jan 29 09:48:00 traceback spew during population Jan 29 09:48:13 thegve: try that Jan 29 09:51:27 good morning everyone Jan 29 10:04:38 Hi all.. are patches to the OE metadata best submitted to the bugtracker? Jan 29 10:04:56 NAbyss: yes Jan 29 11:07:39 JustinP: :) Jan 29 11:16:31 koen|away: any possibility to make planetplanet read blogger beta? Jan 29 11:42:53 mickey|dinner: I added the gnokii bug to oe bugzilla #1826 Jan 29 11:52:29 !oebug 1826 Jan 29 11:52:30 * * Bug 1826, Status: NEW, Created: 2007-01-29 04:16 Jan 29 11:52:31 * * stefan(AT)datenfreihafen.org: CROSS COMPILE Badness in gnokii Jan 29 11:52:32 * * http://bugs.openembedded.org/show_bug.cgi?id=1826 Jan 29 11:55:41 hrw|work: Ah, cute Jan 29 12:15:02 morning Jan 29 12:15:45 If something is doing mkfontscale in pkg_postinst, wouldn't that mean that mkfontscale should be in RDEPENDS? Jan 29 12:16:09 pkg_postinst runs on the target system after the package has been installed right?? Jan 29 12:17:25 tkp: correct Jan 29 12:20:08 NAbyss: thanks Jan 29 12:27:09 I'm having troublee using some of the xorg-font packages Jan 29 12:27:36 I'm trying to add font-util to PACKAGE_INSTALL in my custom image Jan 29 12:27:49 and it complains: Jan 29 12:27:49 Cannot satisfy the following dependencies for font-util: Jan 29 12:27:50 | mkfontscale Jan 29 12:27:57 hi all Jan 29 12:28:06 tkp: Does the mkfontscale package exist? Jan 29 12:28:09 so I tried adding font-util to RDEPENDS and to DEPENDS Jan 29 12:28:13 yes Jan 29 12:28:22 and I can bitbake it byitself fine Jan 29 12:29:17 wait Jan 29 12:29:27 I see the problem I think... Jan 29 12:29:37 mkfontscale package has "inherit native autotools pkgconfig" Jan 29 12:29:42 inherit native? Jan 29 12:29:48 surely that shouldn't be there Jan 29 12:30:03 tht should only be in mkfontscale-native no? Jan 29 12:30:11 yes Jan 29 12:30:12 (which it is also) Jan 29 12:31:31 file a bug? Jan 29 12:33:13 tkp: Yep, with a patch if possible Jan 29 12:33:22 http://bugs.openembedded.org/enter_bug.cgi Jan 29 12:42:57 mindone Jan 29 12:43:00 done* Jan 29 12:50:05 would it be better to maintain a seperate kernel .bb for a board or combine it with the linux-rp kernel? I'm using a lot of the RP patches for my kernel. Jan 29 12:52:19 I'm trying to write a bb file for a package which needs an X11 install to build and to run... Jan 29 12:52:40 I have added virtual/libx11 to DEPENDS, but I think that is not a valid thing to add to RDEPENDS Jan 29 12:52:58 what would I need add to RDEPENDS? Jan 29 12:53:29 Gerrath: Can you remind me which machine this was please :) Jan 29 12:54:15 I don't want to hardcode a specific libx11 package (libx11 or diet-x11) because it should be chosen by setting PREFERRED_PROVIDER Jan 29 12:56:51 tkp: It should autodetect runtime library rdepends Jan 29 12:57:07 RP: oh right Jan 29 12:57:11 (it looks at what your binaries link to) Jan 29 12:57:15 so no need to add to RDEPENDS Jan 29 12:57:19 ok Jan 29 12:57:59 You need to specify something for things like perl scripts but binaries/libraries should "just work" :) Jan 29 12:58:44 that behavior is specific to virtual/libx11? Jan 29 12:58:50 or all virtual/ packages? Jan 29 12:59:15 tkp: No, it works for anything thats dynamically linked Jan 29 12:59:33 oh right. thanks Jan 29 12:59:43 Grep the classes directory for shlibs if you want to see the code that does it Jan 29 13:04:10 RP, LogicPD, I have already modified the linux-rp to do it, I'm just trying to decide if it is the best thing to do. What is the purpose of linux-rp, is it to unify multiple machine kernels? Jan 29 13:11:06 Gerrath: It started as a kernel for the Zaurus devices and has become a little bit broader than that. Its aim is to push things into mainline so I guess it depends what you plan to do with the kernel patches eventually Jan 29 13:27:14 rp, thanks that cleared things up for me :-) Jan 29 13:57:50 morning Jan 29 14:17:40 haha, history repeats http://sourceforge.net/projects/kpdf2/ Jan 29 14:21:03 zecke: qpdf2 again? Jan 29 14:22:43 hrw|work: no kpdf ... Jan 29 14:23:44 chouimat: years ago there was opie-qpdf project which someone forked into qpdf2 Jan 29 14:23:58 chouimat: and qpdf2 is nightmare code Jan 29 14:24:26 hrw|work: I know Jan 29 14:25:22 hrw|work: looks like you can change your flights... I need some minutes... Jan 29 14:31:58 greentux: great Jan 29 14:33:55 hi all Jan 29 14:36:18 hrw|work: which way You have in mind to fly? Jan 29 14:38:38 moment Jan 29 14:40:30 something is wrong with xserver-kdrive bb file somewhere... but I can't sus it out Jan 29 14:40:44 I asked for xserver-kdrive-fbdev to be installed into my image Jan 29 14:40:55 and I end up with xserver-kdrive-fbdev and xserver-kdrive being installed Jan 29 14:41:13 the bb file has "FILES_xserver-kdrive-fbdev = "${bindir}/Xfbdev"" Jan 29 14:41:28 which looks right... that is the only binary I want to be installed Jan 29 14:41:34 but I get all the others too Jan 29 14:42:16 xserver-kdrive is the common files, you need that Jan 29 14:42:30 XorA: but why do I get all the binaries? Jan 29 14:42:47 I don't want Xephyr for example Jan 29 14:43:04 yet I get it even when I ask specifically for xserver-kdrive-fbdev Jan 29 14:44:03 tkp: no idea, doesnt happen here Jan 29 14:45:01 asking for xserver-kdrive-fbdev in my images seems to cause both the xserver-kdrive-fbdev and xserver-kdrive packages to be installed by ipkg Jan 29 14:45:13 tkp: that is correct Jan 29 14:45:37 xserver-kdrive should have one file SecurityPolicy Jan 29 14:45:48 no... Jan 29 14:46:10 looking in install/ I see that file (under lib/), as well as all the binaries (under usr/bin/) Jan 29 14:46:31 tkp: what is in the packages? Jan 29 14:46:49 how do I look inside an ipk? Jan 29 14:46:59 ipkg files blah.ipk Jan 29 14:47:37 tkp: all the Xservers are always built, they are packaged into different packages Jan 29 14:47:52 thats what I thought should happen Jan 29 14:51:06 greentux: lufthansa poz -> muc -> bru and return same way Jan 29 14:51:41 greentux: or pll lot (http://lot.com): poz -> waw -> bru Jan 29 14:52:46 yep. All the binaries are in the main package Jan 29 14:53:22 hrw|work: but lufthansa back at later evening is not cheap Jan 29 14:53:48 greentux: just noticed - they changed since yesterday Jan 29 14:54:04 greentux: I'm checking 23.02 with 26.02 back Jan 29 14:54:13 FILES_${PN} += "${libdir}/xserver/SecurityPolicy" Jan 29 14:54:23 shouldn't that be FILES_${PN} = Jan 29 14:54:32 tkp: no. Jan 29 14:54:44 but thts the only file that should be in there no? Jan 29 14:54:49 tkp: += is proper because you want to ADD this file to normal FILES_{$PN} Jan 29 14:55:10 and what is in the 'normal' FILES_{PN}? Jan 29 14:55:47 tkp: bitbake xserver-kdrive -e|grep ^FILES Jan 29 14:56:38 ILES_bitbake="/usr/bin/* ... Jan 29 14:56:49 hrw|work: ah 26. back... ok Jan 29 14:57:07 /usr/bin is where the unwanted binaries are Jan 29 14:58:21 surely that shouldn't be in FILES_${PN} ? Jan 29 14:58:43 tkp: but FILES_xserver-kdrive-* should have sucked all them away Jan 29 14:59:24 greentux: I assume that your company prefer LH Jan 29 15:00:42 hrw|work: no, every airline... Jan 29 15:00:45 XorA: would that remove them from the work/.../${PN}/install directory? Jan 29 15:00:55 or does that only happen when the ipk is created Jan 29 15:01:18 tkp: yes, files are removed when packaged, unless pb__ update broke that Jan 29 15:01:49 right... well they can't be being removed since they are ending up in my final rootfs Jan 29 15:02:48 I only asked for one thing in my image... xserver-kdrive-fbdev ( a simple test) Jan 29 15:03:26 XorA: what is pb__ update? Jan 29 15:03:48 tkp: pb__ is a person, and he recently made changes to packaging Jan 29 15:03:53 I see Jan 29 15:04:07 but you say this isn't happening your end? Jan 29 15:04:27 tkp: no, but I dont have the 100% upto date code from today Jan 29 15:04:39 pb_: you around? Jan 29 15:04:53 I updated mine on Friday Jan 29 15:05:49 although to be honest, I have always had those unwanted binaries in my final rootfs... it's only now that I have had time to look into it Jan 29 15:05:49 tkp: then I dont think you have those changes then Jan 29 15:06:24 tkp: then I guess something is up with your config or setup as no-one else has complained Jan 29 15:07:39 hmm... can't think for the life of me what that would be Jan 29 15:08:29 ~lart mysql and its fscking binary logs Jan 29 15:08:29 * ibot pries mysql and its fscking binary logs's back open with a screwdriver and flashes a new bootldr to mysql and its fscking binary logs Jan 29 15:36:15 tkp: hello Jan 29 15:38:51 pb_: hi Jan 29 15:39:19 pb_: did you catch my conversation with XorA ? Jan 29 15:39:35 sorry, no Jan 29 15:39:41 nm... Jan 29 15:39:59 I need to go out again in a moment, but feel free to email me if you need something Jan 29 15:40:40 basically, I'm trying to include xserver-kdrive-fbdev in a root image but am finding that I get not only the binary from xserver-kdrive-fbdev, but also those from xserver-kdrive Jan 29 15:40:59 apparently these should have been 'sucked out' by the FILES_... statements Jan 29 15:41:31 they have indeed been copied into the install/ directory of the sub packages, but seem to have been left behind in the main package Jan 29 15:42:07 and I end up with all the files on my rootfs (those from xserver-kdrive-fbdev aswell as all the other ones from all the other sub-packages) Jan 29 15:42:27 XorA: said you were working on the packaging routines... seen this before? Jan 29 15:43:33 re Jan 29 15:45:35 don't think I've seen that before, no Jan 29 15:45:48 I can take a look at it later this evening though Jan 29 15:46:34 pb_: cool... I should be around later so shout me if you need more info Jan 29 15:49:20 righto, will do Jan 29 15:49:27 I'll start a build in the meantime Jan 29 15:56:11 mickey|dinner: Congrats. You managed to rate your interview going before "have you looked at your arse in the mirror." Jan 29 16:15:00 ggilbert: heh. which interview? Jan 29 16:15:21 ah Jan 29 16:15:22 lugradio Jan 29 16:15:22 hehe Jan 29 16:15:44 guess i should download that and blog about it Jan 29 16:15:53 zecke: oh no... kpdf2... Jan 29 16:17:44 mickeyl: :) Jan 29 16:22:17 hmm Jan 29 16:22:17 gosh Jan 29 16:22:24 my accent is awful Jan 29 16:22:39 * mickeyl stops listening to him Jan 29 16:23:38 pb_: are you for hire? I need some articulation lessons and you have the most fantastic accent i have ever heard :) Jan 29 16:24:58 * hvontres|poodle downloads the interview for later amusement Jan 29 16:25:30 mickeyl: how's that book coming along? I'm still available if you need some help with translations :) Jan 29 16:25:38 heh Jan 29 16:25:46 i'm under 200% OpenMoko load and you ask about the book Jan 29 16:25:47 :D Jan 29 16:25:55 seriously though, i have an interesting offer Jan 29 16:25:58 + got Jan 29 16:26:22 OpenSourcePress is interesting to publish it in a dual-track way which means Jan 29 16:26:25 i'll write it in german Jan 29 16:26:31 they care about translation Jan 29 16:26:47 and then partner up with their US division to release it both in german and english Jan 29 16:26:49 * koen|gprs gets flashbacks to 'allo 'allo Jan 29 16:26:55 which sounds about the best option i could ever think of Jan 29 16:27:22 mickeyl: sounds cool... Jan 29 16:27:39 definitly. if it's in german, i can do it in 6 months Jan 29 16:27:45 i have scheduled starting in April Jan 29 16:27:50 for release around christmas Jan 29 16:27:52 * hvontres|poodle wonders if a "Platt Deutsch" edition would work for both places ...:) Jan 29 16:27:57 heh Jan 29 16:28:25 mickeyl: then koen could probably read it too :) Jan 29 16:29:34 * koen|gprs listens to mickeyl Jan 29 16:30:00 edge is fast enough for the lo-fi mp3 :) Jan 29 16:30:04 heh really? Jan 29 16:30:05 impressive Jan 29 16:30:10 ~lart gprs Jan 29 16:30:10 * ibot shoots gprs in the head Jan 29 16:30:50 well, i'll be dardned : http://nds.wikipedia.org/wiki/Linux_op_Platt Jan 29 16:38:53 should conditional patches to a gcc be appended by creating a new bitbake script that appends something to the version? Jan 29 16:39:06 i.e gcc_3.4.3-9x.bb ? Jan 29 16:39:22 where is that mickeyl interview? URL please.. Jan 29 16:39:34 http://uk.lugradio.atchoo.org/season4/lugradio-s04e11-290107-low.mp3 Jan 29 16:39:37 http://www.lugradio.org/ Jan 29 16:40:07 thx Jan 29 16:40:12 will listen at home Jan 29 16:40:30 "then you say 'bitbake something-image'" :) Jan 29 16:40:37 heh Jan 29 16:41:05 was that for me koenjedge? please elaborate I'm confused Jan 29 16:43:08 In general, should toolchain version in bb scripts be considered a way to create alternate patch version of the toolchain? Jan 29 16:43:28 and are non-numberical version numbers acceptable? Jan 29 16:43:39 ....numerical Jan 29 16:43:44 cu Jan 29 16:44:24 nzg: patch to gcc x.y.z need to be in gcc_x.y.z.bb Jan 29 16:45:16 makes sense, but can z be, for example "crunch" or does it need to be numerical Jan 29 16:45:30 if your departing from the standard tool chain Jan 29 16:46:12 why do you *want* a seperate bb if you don't *need* it? Jan 29 16:46:54 So I can leave an intact oe CVS source and create additions in a separate directory Jan 29 16:48:49 or SV synced source as the case may be Jan 29 16:49:20 we discussed this on a bit Friday Jan 29 16:49:30 cvs? Jan 29 16:49:33 cvs?!?! Jan 29 16:49:35 cvs?!?!?!?!? Jan 29 16:49:38 ok Jan 29 16:49:41 that was a typo Jan 29 16:49:48 or a braino anyway Jan 29 16:50:22 But it was suggested to append the patches with SRC_URI+= which makes sense Jan 29 16:50:29 but where do I append? Jan 29 16:50:46 clearly I want to append to the gcc_version for all targets Jan 29 16:51:03 inside gcc_x.y.z.bb Jan 29 16:51:13 where z is a new creation? Jan 29 16:51:17 no Jan 29 16:51:24 don't create new .bbs Jan 29 16:51:41 ah Jan 29 16:51:42 look at our .inc files Jan 29 16:51:47 edit existing ones and send patches to us Jan 29 16:51:48 maybe one of those fits Jan 29 16:51:55 yeah, or that Jan 29 16:52:00 I can Jan 29 16:52:04 * koen|edge want's better crunch support in 4.1.x Jan 29 16:52:13 s/want's/wants/ Jan 29 16:52:46 but my concern is that adding conditional appends directly to the existing bb's is going to create an ever growing pile of nastyness Jan 29 16:52:55 like C code filled with ifdefs Jan 29 16:53:28 I have not looked at the inc files though, so I'll take a break from talking and do some readin Jan 29 16:59:33 the .inc's are shedding little light Jan 29 16:59:37 do you know of an example Jan 29 16:59:53 where a patch is conditionally applied to the a specific toolchain version? Jan 29 17:00:53 yes Jan 29 17:00:59 inside gcc_x.y.z.bb Jan 29 17:01:06 * koen|edge feels like repeating himself Jan 29 17:01:49 I know, but perhaps I'm missing something Jan 29 17:01:53 for instance Jan 29 17:02:03 If I look in gcc_3.4.3.bb Jan 29 17:02:16 I see a lot of SRC_URI's applied to make that version Jan 29 17:02:20 gcc3 is dead Jan 29 17:02:41 I think that's a bit premature Jan 29 17:02:50 I know it would be nice if it was Jan 29 17:03:11 let me give you a little background to my motivation Jan 29 17:03:36 The Cirrus ep93xx series Jan 29 17:03:53 Has a lot of stable working patches available for it's toochain Jan 29 17:04:04 heh Jan 29 17:04:05 It also has a good, fully supported kernel Jan 29 17:04:09 heh Jan 29 17:04:13 I disagree Jan 29 17:04:22 cirrus has a bunch of ugly hacks Jan 29 17:04:27 perhaps so Jan 29 17:04:37 but if I want working crunch support Jan 29 17:04:45 and I don't want to write it from scratch Jan 29 17:04:50 what options do I have? Jan 29 17:05:13 get the latest kernel from kernel.org, that has good crunch and ep93xx support Jan 29 17:05:22 then build yourself an eabi system Jan 29 17:05:38 then expand the crunch support in gcc 4 Jan 29 17:05:49 gcc4 has crunch support, but is missing some errata Jan 29 17:06:00 that sounds dangerous Jan 29 17:06:00 * koen|edge has a ep9312 board on his desk Jan 29 17:06:21 I'm working with a ep9302A derivative Jan 29 17:06:45 it isn't dangerous, it just means some apps won't work with crunch Jan 29 17:07:00 which isn't a big loss, since you don't want to build every app with crunch support Jan 29 17:07:00 will they work with the older toolchain from cirrus? Jan 29 17:07:10 o Jan 29 17:07:37 still I will need to quantify this, as this will become a commercial system Jan 29 17:08:39 hence I was hoping to use the crunch patches released by cirrus Jan 29 17:08:39 then at least I have a finger to point ;-) Jan 29 17:08:39 your best bet is to hire Lennert Buythenhek :) Jan 29 17:08:41 lol Jan 29 17:09:30 well I was thinking integrating the existing crunch patches from cirrus wouldn't be too difficult, my main problem so far is purely academic Jan 29 17:09:46 I.E. where do I put these patches in that's not a hack Jan 29 17:10:02 even if, granted, the patches themselves are hacks Jan 29 17:10:18 in gcc_3.4.b.bb Jan 29 17:10:24 s/b/6/ Jan 29 17:11:13 do what now Jan 29 17:11:25 is b an invented number Jan 29 17:11:37 or am I missing something Jan 29 17:16:39 nzg: just read your article about eclipse in "Embedded Computing Design" Jan 29 17:16:52 oh yea? Jan 29 17:16:55 nzg: do you OE much at EMAC? Jan 29 17:17:06 Just began Jan 29 17:17:17 That article was pre-oe Jan 29 17:17:52 nzg: good stuff. I just built an OE distro for a customer using one of your boards. Jan 29 17:18:13 beautiful :-) Jan 29 17:18:18 which one? Jan 29 17:19:00 nzg: it was a pentium board -- lets see ... Jan 29 17:20:31 nzg: PCM-9371 -- I think it came from EMAC, but not absolutely sure. Jan 29 17:20:47 well sell that one yes Jan 29 17:21:02 nzg: we were using a Geode board in previous versions Jan 29 17:21:36 Yes the geode appears to have "evolved" into something that makes less sense Jan 29 17:21:44 nzg: http://bec-systems.com/web/content/view/57/9/ Jan 29 17:21:51 We going towards the arm Jan 29 17:21:52 nzg: heh Jan 29 17:22:48 nzg: yeah, ARM is good for many applications -- most of what I do. Jan 29 17:23:28 it's the future.. if I can just resolve this Cirrus crunch patch thing Jan 29 17:27:48 nzg: so is the issue the patches themselves, or figuring out how to get them into an OE toolchain? Jan 29 17:50:23 the latter Jan 29 17:50:47 as discussed earlier, the place to include them in the SRC_URI is questionable Jan 29 17:52:04 I think I understand koen|edge's original comment about putting them directly into the gcc_ bb script Jan 29 17:52:18 but I fear they make break other things, so their compilation should be conditional Jan 29 17:52:55 Also, as mentioned, the cirrus patches seem to be held in rather low regard so I doubt they would stay in the main source even if I released the patch Jan 29 17:53:04 nzg: nod Jan 29 17:53:39 At the moment I'm going to put them into a new bb script in a separate directly included by the path Jan 29 17:53:57 with some string appended to the version Jan 29 17:54:06 nzg: yeah, that makes sense until you get everything working. Jan 29 17:54:08 but I'm open to suggestions as to a cleaner way to do this Jan 29 17:54:35 nzg: have you looked at bbcollections -- will probably eventually come in handy. Jan 29 17:54:52 I have not, that's the first I've heard of it, please elaborate Jan 29 17:56:35 nzg: http://bitbake.berlios.de/manual/ch04s02.html#id870544 Jan 29 17:58:10 that might be handy, thx Jan 29 17:58:52 nzg: real handy for maintaining custom packages. For most of my stuff, I just maintain two trees. The pristine oe.dev tree and a custom tree. As I push stuff into the oe repo, I take it out of the custom tree. Jan 29 18:00:19 That's been my strategy so far, but I've been renaming everything as I was unaware of bbcollection, this does simplify things Jan 29 18:00:52 mickeyl: your accent isn't that bad. :) Jan 29 18:02:06 :) Jan 29 18:02:07 thank Jan 29 18:02:09 s Jan 29 18:14:01 koen|edge_: you've probably already seen this: http://lists.freedesktop.org/archives/cairo/2007-January/009368.html Jan 29 18:14:35 hi, all! Jan 29 18:15:47 if I set package system to dpkg, will I be able to use these with normal dpkg-buildpackage tools or it will be made in hacky way? Jan 29 18:16:35 so I want to build basic image using OE and use it as scratchbox rootstrap to build some debian packages Jan 29 18:16:49 will it be possible? Jan 29 18:19:20 mreimer: I pasted thatin #angstrom a few minutes ago :) Jan 29 18:19:28 koen|edge_: :-) Jan 29 18:19:33 koen|edge_: I figured Jan 29 18:26:15 hi Jan 29 18:31:39 i'm getting creazy with this ... configure: error: Couldn't find Xaw library in native compilation... guys.. any idea ? I'm try to compile ng-spice on openzaurus 5.4.1 ... native Jan 29 19:44:32 any PXA27x gurus here? Jan 29 19:45:22 is it possible to stop resume process in kernel and continue sleeping on some condition? Jan 29 19:47:03 hmh, I'd like to do something like this, too. But it doesn't look like there is a mechanism in the kernel for this. Jan 29 19:47:36 doesn't dyntick do exactly that? Jan 29 19:47:40 maybe you can change the cpu clock Jan 29 19:48:17 I thought dyntick just sleeps, does it suspend? Jan 29 19:48:57 ah, right, it just sleeps Jan 29 19:49:08 slapin_nb: sort of; see arch/arm/sharpsl_pm.c:corgi_pxa_pm_enter() Jan 29 19:50:10 slapin_nb: basically it decides whether to do a low-level cpu sleep again after wake up, or return from cpu sleep Jan 29 19:51:11 mreimer: this is not in the upstream kernel yet? Jan 29 19:51:39 pH5: looks like it's in 2.6.19 Jan 29 19:52:20 pH5, slapin_nb: ah, the code's been rearranged a bit, but it is there in 2.6.19 Jan 29 19:52:40 pH5, slapin_nb: the key is the pxa_pm_enter() wrapper Jan 29 19:53:21 mreimer, which branch it is in? Jan 29 19:53:41 i.e. which kernel should I check out? Jan 29 19:53:45 slapin_nb: I was just looking in a vanilla 2.6.19; don't know about hh.org Jan 29 19:56:00 mreimer: so the secret is to replace pxa_pm_enter with our own version via pm_set_ops like it is done in arch/arm/common/sharpsl_pm.c? Jan 29 19:56:43 pH5: yes. for a simpler example see h2200_pm.c Jan 29 19:57:08 pH5: I'm guessing you want to do this to update LEDs after full charge while suspended? Jan 29 19:58:04 pH5: h2200_pm doesn't re-sleep; what you would do is after calling pxa_pm_enter_orig(state), decide whether to sleep again, and if so, call pxa_pm_enter_orig(state) again Jan 29 19:58:15 mreimer: that, and I want to wake up on the CPLD irq (to catch bt interrupts), but not when the AC cable is pulled, for example. Jan 29 19:58:39 pH5: that should be easy Jan 29 19:59:10 pH5: actually, you'll probably want to do a partial wake on AC out so you can update the leds. You should be able to do that too. Jan 29 20:00:31 thanks, all! Jan 29 20:00:35 mreimer: right. thanks for the pointer! Jan 29 20:01:59 pH5: the problem I've stalled on for ds2760 users is how to check battery status while mostly suspended (e.g. right after cpu has awakened, but before returning from pxa_pm_enter(). It's probably not safe to call into the w1 system to get ds2760 status because it's probably frozen. Jan 29 20:04:57 mreimer: would it be possible to selectively wake up the w1 subsystem? Jan 29 20:05:40 this goes through so many layers of abstraction :( Jan 29 20:06:48 I have a really strange problem with one of my bb files... it's dyeing with this: http://rafb.net/p/Ndjb9792.html Jan 29 20:07:01 pH5: I don't know. I should ask the w1 maintainer. I doubt it. One possibility I thought of is to add a function to ds1wm.c that we call when the system is suspended; since the driver knows it is suspended it could do it a different way. At least the code would all be in the same place then, but it would be a bit messy. Jan 29 20:07:09 odd thing is, if I simply rename the bbfile and rebake then all is fine Jan 29 20:07:19 I have tried -c clean on it but I still get the problem Jan 29 20:07:33 I also tried removing the cache file... problem still exists Jan 29 20:23:25 mreimer: I think a single function that can be called by all ds1wm+ds2760 devices would be a good thing for now. Jan 29 20:23:34 pH5: agreed Jan 29 20:24:52 I like the should_wakeup function in the zaurus code. I think I'll try to copy that structure (or the parts that I need). Jan 29 20:28:58 Who actually maintains ipkg? Jan 29 20:38:31 I'm having an oe build problem Jan 29 20:38:49 it's complaining that quilt is being used wrong Jan 29 20:38:54 or appears to be Jan 29 20:39:04 i.e. ERROR: Exception:__builtin__.CmdError Message:Command Error: exit status: 1 Output: Jan 29 20:39:04 Usage: quilt [--trace[=verbose]] [--quiltrc=XX] command [-h] ... Jan 29 20:39:05 quilt --version Jan 29 20:39:21 has anyone else seen this? any ideas? Jan 29 20:39:56 do I need to do some sort of clean before changing the version of the toolchain I'm building? Jan 29 20:41:17 nzg: did you move your dirs around? Jan 29 20:41:35 I don't believe so Jan 29 20:42:06 I did add a bbcollections to local.conf Jan 29 20:42:31 one of my packages contains a load of shared libraries, but when it gets installed on the system, I get two copies of each - the second lot in a .debug directory... Jan 29 20:42:47 how do I get these to go in to ${PN}-debug? Jan 29 20:43:08 03kristoffer 07org.oe.dev * ref3ac8e2... 10/ (4 files in 3 dirs): Jan 29 20:43:08 ipkg/ipkg-0.99.163/patch-ipkg-localfilename : Addition of patch Jan 29 20:43:08 * This patch fixes one of many bugs in ipkg, involving Jan 29 20:43:08 localfilename. Submitted by Alexandre Orlov Jan 29 20:43:08 from the Jlime dev team. Confirmed working. Jan 29 20:43:09 * Applied patch to ipkg_0.99.163.bb Jan 29 20:43:19 tkp: You probably need to set FILES_${PN}-dbg to catch them properly Jan 29 20:43:41 what would I set it to Jan 29 20:44:00 tkp: Have a look at the defaults in bitbake.conf Jan 29 20:44:06 RP: thanks Jan 29 20:45:08 ah ok I get it... it's because I'm installing the libs into a non standard location Jan 29 20:51:14 letting bb start for scratch, seeing if that fixes the quilt problem Jan 29 21:03:49 Kristoffer: did you send this patch upstream? Jan 29 21:06:20 zecke, not yet, just got my hands on it. But will do Jan 29 21:07:16 Kristoffer: make really sure you do. Actually this patch should have been sent upstream before adding it to OE Jan 29 21:08:34 zecke, Oki, will remember that. You got an email btw? Looking at the http://handhelds.org/moin/moin.cgi/HandheldsPeople but cant see a ipkg maintainer there Jan 29 21:09:07 ~seen pigi Jan 29 21:09:12 pigi was last seen on IRC in channel #asterisk, 244d 7h 5m ago, saying: 'is there anyone that has successfully got asterisk working with eicon diva pci cards (passive cards ) ?'. Jan 29 21:09:26 hmm, bad sign :) Jan 29 21:09:30 :( Jan 29 21:09:46 ping hrw|gone Jan 29 21:11:01 wow Jan 29 21:13:18 greentux: 5 minutes Jan 29 21:13:30 greentux: have to check few mails Jan 29 21:13:32 evening guys Jan 29 21:13:49 hi hrw Jan 29 21:15:25 ~botmail for Kristoffer Erik Hovland looks like a good bet for the patches Jan 29 21:18:24 hrw: I have 20 minutes left ;) Jan 29 21:18:29 hrw: joke Jan 29 21:19:55 ;D Jan 29 21:33:18 greentux: now you are gone? Jan 29 21:33:35 hrw: no Jan 29 21:35:18 03pH5 07org.oe.dev * r3ce85c1d... 10/ (7 files in 3 dirs): bluez: add bluez-libs and bluez-utils 3.9 Jan 29 21:49:47 hrw slept down... Jan 29 21:50:16 or lagged Jan 29 21:51:06 XorA & pb__ : re my problem with xorg-xserver-kdrive not splitting up the PACKAGES correctly... Jan 29 21:51:43 tkp: yup Jan 29 21:51:45 hrw ok we can discuss that also tomorrow Jan 29 21:52:04 it does. It's just that the bb file only defines 6 different xserver packages (Xfbdev, Xfake, Xephyr, Xepson, Xsdl & Xw100) Jan 29 21:52:06 greentux: I know Jan 29 21:52:38 so things like Xvesa nd Xvia amongst others end up in the default package... which also getss installed when you ask for one of the subpackages Jan 29 21:53:01 tkp: fancy putting a patch on bugtracker> Jan 29 21:53:03 ? Jan 29 21:53:06 which I would think is the wrong behaviour Jan 29 21:53:10 XorA: suure Jan 29 21:53:29 just not too sure how much to split things up... Jan 29 21:53:46 I'm not too familiar with all tyhe related xserver binaries Jan 29 21:54:47 Xvia and Xvesa should presumably go into xorg-xserver-xvia/xvesa Jan 29 21:55:10 or rather... xorg-xserver-kdrive-xvia/xvesa Jan 29 21:55:58 tkp: yes, sounds good Jan 29 21:56:13 tkp: though I wonder why they suddenly startted building Jan 29 21:56:20 what about card specific ones... like Xati & Xnvidia Jan 29 21:56:27 tkp: you building for x86? Jan 29 21:56:30 yes Jan 29 21:57:21 tkp: Id put them all in their own package Jan 29 21:57:35 right... makes sence Jan 29 21:57:55 your only ever going to want one of them... whichever one is most suitable for your setup Jan 29 21:58:01 tkp: only thing I dont know is what happens on other platforms when they dont exist. But we fix that when we test Jan 29 21:58:15 point Jan 29 21:58:34 hrw: ok. give me a sign when I can do something for you. going to bed now... Jan 29 21:58:50 * XorA sends greentux a For Sale Jan 29 21:58:57 greentux: sweet dreams and thx Jan 29 21:59:23 XorA: ? Jan 29 21:59:34 greentux: a sign :-) Jan 29 22:00:07 XorA: yeah will improve my english this year. perhaps malta ,perhaps england... 2 weeks intensive Jan 29 22:00:30 greentux: we dont do English well :-) Jan 29 22:01:16 XorA: you are in england or scotland? Jan 29 22:01:21 XorA: the maltese do... Jan 29 22:02:06 hrw: Scotland, but I am English Jan 29 22:02:59 hrw: fine, You no what I can organize. Now its to You to shake the flight plans :) Jan 29 22:04:13 I know,, I know Jan 29 22:05:18 I just have to find how to pass my mental block for getting such help for free. Jan 29 22:06:10 ? Jan 29 22:06:21 XorA: just saw http://www.handhelds.org/hypermail/oe-commits/40/4016.html ... so this was already done at one point? Jan 29 22:06:48 greentux_away: will talk about it tomorrow Jan 29 22:07:00 or at least, it had been split up more than it is now Jan 29 22:07:24 hrw: so trying to build OZ 3.5.4.2-rc2 for poodle on my machine is a disaster Jan 29 22:09:12 * tkp wonders why it was changed Jan 29 22:10:23 tkp: someone was probably only building arm when we moved to xorg 7.0 and it got forgotten Jan 29 22:10:50 tkp: but at least a patch in bugzilla will remind people Jan 29 22:11:09 tkp: and if there is a sinister reason in the cupbaord it might show its head :-D Jan 29 22:11:18 XorA: which bitbake are you using? my setup was working last week with 1.6.3 Jan 29 22:11:32 ;) just sorting it. I'll have a patch up in a bit. Jan 29 22:11:39 hvontres|poodle: its the amd64 bit I think is causing problems Jan 29 22:11:52 hvontres|poodle: .dev works no probs Jan 29 22:12:09 XorA: yeah.. I definitely don't have that problem...dual P-II 333 :( Jan 29 22:14:25 ~seen koen Jan 29 22:14:54 mickeyl: that patch is monstrous in its speed increase Jan 29 22:14:59 koen was last seen on IRC in channel #openmoko, 1d 13h 15m 4s ago, saying: 'I didn't even know about SITEINFO_ENDIANESS :)'. Jan 29 22:15:17 XorA: really? wooot! Jan 29 22:15:44 everything that surpasses 2.6 would be _very_ welcome Jan 29 22:15:54 by the way Jan 29 22:16:08 mickeyl: I shall check in the .bb and you can give it a spin Jan 29 22:16:09 ~pester pb_ for adjusting integer-pixops patch to 2.6.10 and 2.10.x Jan 29 22:16:14 XorA: yeah Jan 29 22:16:16 pb_ for adjusting integer-pixops patch to 2.6.10 and 2.10.x: Are we there yet? .. Are we there yet? .. Are we there yet? see http://beverlys.net/LJ/BuggingYou.swf Jan 29 22:16:17 mickeyl: looked at http://ewi546.ewi.utwente.nl/tmp/hrw/t/030_CVE-2007-0010_error-handling-in-pixbuf-loaders.patch one? Jan 29 22:16:31 hrw: i looked, but I'm afraid I don't get it :) what does it do? Jan 29 22:16:44 mickeyl: google for 'cve 2007 0010' Jan 29 22:16:50 k Jan 29 22:16:51 security for broken images Jan 29 22:17:20 hi ! Jan 29 22:17:50 ah that one Jan 29 22:17:56 ok, let me review again Jan 29 22:18:46 i've got a problem building a package i'm trying to add to existing ones; it does `make' properly but the package is empty, I don't understand what's wrong :( Jan 29 22:19:09 vinvin: autotooled package? Jan 29 22:19:10 output is http://vinvin.dyndns.org/~vinvin/output.txt Jan 29 22:19:12 no Jan 29 22:19:13 hrw: i think we can apply that Jan 29 22:19:18 hand made hrw Jan 29 22:19:35 http://vinvin.dyndns.org/~vinvin/light_1.0.bb for the bb file Jan 29 22:20:27 vinvin: install -d ${D}/usr/light ${D}/usr/light/www Jan 29 22:20:30 its ugly Jan 29 22:20:50 to put two on the same line ? Jan 29 22:20:50 03xora 07org.oe.dev * r51d3ccc9... 10/ (3 files in 3 dirs): Jan 29 22:20:50 cairo_1.3.12.bb : add patch from jorn at openedhand.com that promises a Jan 29 22:20:50 12% speedup of cairo. Jan 29 22:20:52 ${D}${bindir} ${datadir}/light/www/ -> /usr/bin and /usr/share/light/www/ Jan 29 22:21:12 oh that Jan 29 22:21:15 vinvin: be FHS friendly or we will reject it Jan 29 22:21:28 mickeyl: let the benchmarking begin Jan 29 22:21:56 vinvin: then it will get packaged Jan 29 22:21:58 it's not meant to be added to the real project, but to make this software work embedded hrw Jan 29 22:22:05 XorA: excellent - thanks Jan 29 22:22:27 the app looks for www where it is launched it's just why i put it at the same place Jan 29 22:23:20 vinvin: make it FHS friendly - it does not hurt and will make things easier Jan 29 22:23:47 allright... Jan 29 22:23:47 vinvin: FILES_${PN} = "/usr/light/*" will probably help you but package is ugly Jan 29 22:24:35 the real question for me was why does "the following files were installed but not shipped in any package" ? Jan 29 22:24:57 i'll recode it to look for /usr/share/light/www but that doesn't solve the problem Jan 29 22:24:58 vinvin: bitbake -b light_1.0.bb -e|grep ^FILES Jan 29 22:25:29 vinvin: default FILES contain FHS compatible directories Jan 29 22:25:38 oh ok Jan 29 22:26:59 damned i've forgotten the password to ssh it Jan 29 22:29:15 allright hrw, so if i just move the files to dirs in FILES_light they will be packages if i follow you well ? Jan 29 22:30:43 yes Jan 29 22:30:50 thank you very much ! Jan 29 22:32:18 np Jan 29 22:36:34 mickeyl: have a minute or few? Jan 29 22:37:00 hi Rod Jan 29 22:37:27 rwhitby: is it hard to make fatslug? Jan 29 22:37:42 hrw: hardware or software -wise? Jan 29 22:38:18 both. for example: I would buy nslu2 and want to make it 266MHz (like new one are) and 128/256MB ram Jan 29 22:39:49 hrw: one minute :) Jan 29 22:40:18 hrw: hardware is too hard for me personally, but I know of about three people who have done it. Jan 29 22:40:33 software should be a breeze for you - just run Apex and change the cmdline in the kernel. Jan 29 22:40:51 rwhitby: my friend has proper equipment and knowledge in soldering smd stuff Jan 29 22:41:19 one even work in company with hardcore smd soldering equipment Jan 29 22:42:08 hrw: should be doable then. I've never tried it myself. Jan 29 22:44:48 Patch for xserver-kdrive is on bugtracker Jan 29 22:44:59 thx rwhitby Jan 29 22:45:00 this saved over 10mb on my final image! Jan 29 22:45:54 * XorA|gone wants a fat slug too then Jan 29 22:46:44 tkp: looks OK. need PR bump added only Jan 29 22:46:52 doh! Jan 29 22:47:41 mickeyl: I sent you few messages. feel free to answer in some time Jan 29 22:47:43 * tkp updated the patch Jan 29 22:48:35 tkp: you are one of those guys which do not use realnames in emails? Jan 29 22:48:49 oki Jan 29 22:48:53 tomorrow morning Jan 29 22:48:56 going to sleep now Jan 29 22:48:56 really? Jan 29 22:48:57 mickeyl: ok. Jan 29 22:48:58 gÄnight folks Jan 29 22:49:02 what name do I use?\ Jan 29 22:49:03 cu mickeyl Jan 29 22:49:14 Reporter: STV Jan 29 22:49:59 ah right/// coz there are a couple of us from stv that use the same account Jan 29 22:50:04 ah. Jan 29 22:50:38 what is the right way to do the make install in bb files ? Jan 29 22:51:07 do_install() { oe_runmake install DESTDIR="${D}"} Jan 29 22:51:13 but it is default one iirc Jan 29 22:51:56 ok :) Jan 29 22:54:31 tkp: Tom Kirkpatrick is realname? Jan 29 22:54:39 indeed ;) Jan 29 22:55:24 then look at CIA-19 Jan 29 22:56:41 CIA-19? Jan 29 22:57:04 CIA notify #oe about commits Jan 29 22:58:52 should I append something to default compile options or define do_compile to do it just the way it is defined in the makefile ? Jan 29 22:59:28 vinvin: bitbake -b light_1.0.bb -e|less Jan 29 23:01:32 hrw the thing is I tried to remove my little rules to see, and i don't have -I/usr/include/upnp anymore Jan 29 23:01:42 so it does not compile Jan 29 23:01:58 vinvin: EXTRA_CFLAGS = "-I{$includedir}/upnp" Jan 29 23:02:07 When I add a patch to a bb file, the cache doesn't appear to be noticing on the next build run, is this correct? Jan 29 23:02:08 cool Jan 29 23:02:10 thanks Jan 29 23:02:22 do I have to delete the cache to try again? Jan 29 23:02:23 nzg: it should print "1 parsed" Jan 29 23:02:48 hrw:and then rebuild it? Jan 29 23:03:00 nzg: yes Jan 29 23:03:20 hrw:hmm, doesn't look like it is, I'll try to get some more data Jan 29 23:03:29 nzg: adding patch should usually also bump PR if result affect packaging Jan 29 23:03:36 or build results Jan 29 23:03:51 hrw that does not work, should I just put that in the bb file ? or do I have to define do_compile() something Jan 29 23:03:52 ? Jan 29 23:04:18 moment Jan 29 23:04:22 hrw:but what about in development? Jan 29 23:04:33 hrw:when the PR isn't getting bumped? Jan 29 23:04:48 on every change Jan 29 23:05:12 I delete the ipk file manully Jan 29 23:05:18 nzg: I usually add .X - PR = "r17.0" r17.1 etc until it will work Jan 29 23:05:33 vinvin: define do_compile() Jan 29 23:05:46 empty ? Jan 29 23:06:06 kk Jan 29 23:06:25 vinvin: look at irda-utils/irdadump_0.9.16.bb Jan 29 23:06:32 ok thanks Jan 29 23:07:00 nzg: or do 'bitbake -crebuild yourpackage' Jan 29 23:07:22 oh yeah ok Jan 29 23:08:56 cu guys - I go sleep now Jan 29 23:09:05 its 00:42 Jan 29 23:09:33 :( Jan 29 23:09:39 good night hrw|gone thanks a lot Jan 29 23:09:48 np Jan 29 23:10:15 CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include/upnp Jan 29 23:10:18 what ? o_O Jan 29 23:10:23 greentux_away: sorry.. I forgot to identify to nickserv so you was not getting my messages. Jan 29 23:10:54 vinvin: use EXTRA_CFLAGS += "-I${STAGING_INCDIR}/upnp/" and make sure that upnp stage headers/libs there Jan 29 23:11:13 o_O ok Jan 29 23:11:20 vinvin: you are CROSS compiling - /usr/lib /usr/include ARE not good Jan 29 23:11:28 cu again Jan 29 23:11:31 okay Jan 29 23:11:34 cu :) Jan 29 23:17:25 hrw|gone: i didn't get any messages either. please repost Jan 29 23:17:31 g'night (now really) Jan 29 23:22:11 i guess putting the install commands in the makefile wasn't a good idea since it doesn't know {D} Jan 29 23:25:28 i hope i did not miss an answer :) Jan 29 23:39:30 vinvin: you didn't Jan 29 23:40:20 :) Jan 30 00:02:54 I'm using a pre-built toolchain so have to have manually include a load of shared libs into my image Jan 30 00:03:16 is there some way I can get a list of all shared libs that are missing? Jan 30 00:11:41 find all the executables in your image, run ldd on them, and compare the resulting lists to the shared libraries you do have Jan 30 01:42:45 hi everyone, newbie, looking for an easy answer: directfb is failing. it looks like bitbake is trying to make rc2, but only rc3 is posted at directfb.org Jan 30 01:44:28 SHELL: Building directfb Jan 30 01:44:28 NOTE: preferred version 2.5 of glibc not available Jan 30 01:44:28 NOTE: package directfb-0.9.99+1.0.0rc2: started Jan 30 01:44:28 NOTE: package directfb-0.9.99+1.0.0rc2-r1: task do_fetch: started Jan 30 01:44:28 NOTE: fetch http://directfb.org/downloads/Core/DirectFB-1.0.0-rc2.tar.gz Jan 30 01:44:29 --10:56:14-- http://directfb.org/downloads/Core/DirectFB-1.0.0-rc2.tar.gz Jan 30 01:44:31 => `/home/paul/sources/DirectFB-1.0.0-rc2.tar.gz' Jan 30 01:44:33 Resolving directfb.org... 212.227.87.76 Jan 30 01:44:35 Connecting to directfb.org|212.227.87.76|:80... connected. Jan 30 01:44:37 HTTP request sent, awaiting response... 404 Not Found Jan 30 01:44:39 10:56:15 ERROR 404: Not Found. Jan 30 01:44:41 NOTE: Task failed: Fetch failed: http://directfb.org/downloads/Core/DirectFB-1.0.0-rc2.tar.gz Jan 30 01:44:43 NOTE: package directfb-0.9.99+1.0.0rc2-r1: task do_fetch: failed Jan 30 01:44:45 ERROR: TaskFailed event exception, aborting Jan 30 01:44:47 NOTE: package directfb-0.9.99+1.0.0rc2: failed Jan 30 01:44:49 ERROR: Couldn't build 'directfb' Jan 30 01:44:51 BB>> Jan 30 01:44:53 oops sorry guys Jan 30 01:46:19 ~pastebin Jan 30 01:46:20 i heard pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.ca, or http://channels.debian.net/paste, or http://rafb.net/paste/ Jan 30 01:46:45 thanks, i'll look for it Jan 30 01:46:51 please create a bug Jan 30 01:47:27 i thought since rc3 was just released, something needs to be updated. I can fix if someone can point me in the right direction... Jan 30 01:49:06 you can alsways try renaming the bb file and see if it pulls and builds the right one Jan 30 01:53:09 ok, changed the name and the URL contained within to reflect the new version. File downloaded and now it's complaining about failed patch... Jan 30 01:58:31 looks like it's trying to patch an outdated version, which makes sense. How might I go about created new patches for the new version of directfb? **** ENDING LOGGING AT Tue Jan 30 02:59:57 2007