**** BEGIN LOGGING AT Sat Jan 05 02:59:57 2008 Jan 05 10:39:53 03pfalcon 07org.oe.dev * r153dd990... 10/ (5 files in 3 dirs): gnome-mplayer: Fix segfaults decoding non-utf8 tags in media. Jan 05 10:39:57 03pfalcon 07org.oe.dev * rfc095692... 10/ (4 files in 2 dirs): Jan 05 10:39:57 xcursor-watchonly-theme: Variation of xcursor-transparent-theme, leaving watch cursor alone. Jan 05 10:39:57 * So there was a visual feedback to the user on long-running operations. Jan 05 10:40:03 03pfalcon 07org.oe.dev * r304f7e47... 10/ (2 files in 2 dirs): xcursor-watchonly-theme 0.1.1: Fully replaces xcursor-transparent-theme. Jan 05 10:41:40 morning all Jan 05 10:43:58 morning Jan 05 10:44:21 hi morning Jan 05 10:44:26 :) Jan 05 10:44:45 good morning CoreDump, RP Jan 05 13:26:15 hi all, can some one help me with bugs no 3522 , 3594 and 3572 (scummvm does not compile) ? I have included a patch to bug 3522 which should similar solve the others as well. Jan 05 13:26:36 s/compile/configure Jan 05 13:56:58 03coredump2 07org.oe.dev * r32cafe5c... 10/ (4 files in 3 dirs): openmoko-gps: Update to SVNR9 Jan 05 14:00:10 still stuck generating locales for the simpad... And now it can't be apparmour, as I am building under the same root as my working akita builds Jan 05 15:03:15 hi everybody Jan 05 16:01:32 re Jan 05 16:29:09 03mickeyl 07org.oe.dev * r7a3bad45... 10/ (4 files in 4 dirs): add python-gsmd, python bindings for the openmoko libgsmd Jan 05 16:56:24 hey all, if I have a bbclass, modeled on angstrom-mirrors.bbclass, but with PREMIRROR instead of MIRROR, is there any reason it wouldn't try the PREMIRROR first when fetching an http source? Jan 05 16:56:43 03coredump2 07org.oe.dev * re244d1fc... 10/ (1 packages/espeak/espeak_1.30.bb): espeak: Add staging Jan 05 16:56:50 03coredump2 07org.oe.dev * rba35013b... 10/ (5 files in 3 dirs): Jan 05 16:56:50 speech-dispatcher: Jan 05 16:56:50 - Fix RPROVIDES so it doesn't pull in a ton of -dev packages Jan 05 16:56:50 - Update EXTRA_OECONF to exclude modules we do not need and enable espeak Jan 05 16:56:50 - Update configs to set espeak module as default and fix logfile destination Jan 05 16:56:51 - Install initscript (this could use some work...) Jan 05 16:56:54 - Tested to work on fic-gta01 Jan 05 16:56:57 03coredump2 07org.oe.dev * ra97c52f2... 10/ (3 files in 3 dirs): speech-dispatcher: Fix packaging of configfiles Jan 05 17:34:41 does OE support targets with samsung s32c2510-based arm cpus? Jan 05 17:39:29 while i don't know off-hand if there's support for a specific ref board already Jan 05 17:39:34 it's just another arm board, so yes. Jan 05 17:40:02 actually, it's a router :0 Jan 05 17:40:18 one of the 2mb-flash Jan 05 17:42:52 Ah, if you can update your specific router, I don't know, sorry Jan 05 17:43:00 I'm used to dealing with pre-production HW :) Jan 05 18:18:47 03coredump2 07org.oe.dev * r897413a9... 10/ (5 files in 3 dirs): navit: Update .inc, include navit.launcher to enable the low-mem workaround on-demand. Jan 05 19:32:52 03mickeyl 07org.oe.dev * r5e5ed727... 10/ (1 packages/u-boot/u-boot-openmoko_svn.bb): uboot-openmoko svn disable eabi fix from rev. 3774 onwards, it's in OM svn now Jan 05 19:44:25 Unknown option: -n Jan 05 19:44:32 from python during image creation Jan 05 19:44:39 I got this in an email ... Jan 05 19:44:42 any thoughts? Jan 05 19:45:23 Crofton: -n for what script :) Jan 05 19:48:45 re Jan 05 19:49:08 console-image.bb? Jan 05 19:49:30 Crofton: that is not a script, but that is elitist? Jan 05 19:49:33 I'm looking at messages that were emailed to me by a guy trying to build an image for the omap2430sdp ... Jan 05 19:50:12 I am going to pastebin the while thing Jan 05 19:50:23 I was hoping it was an "obvious problem" Jan 05 19:50:40 http://www.pastebin.org/14142 Jan 05 19:50:41 Crofton: did he mail you the full script that is executed to build the image? Jan 05 19:50:50 no, just a messgae Jan 05 19:51:03 image.bbclass:99: ${PYTHON} `which bbimage` -n "${IMAGE_NAME}" -t "$type" -e "${FILE}" Jan 05 19:51:52 Crofton: there are three -n in the image.bbclass. ask him to send you the script that bitbake is executed. run.do_rootfs or such Jan 05 19:52:13 I hope he shows up here :) Jan 05 19:53:30 looks like do_rootfs failed Jan 05 19:53:35 I gotta run bbl Jan 05 20:04:09 ~lart feature creep Jan 05 20:04:09 * ibot puts on a hockey mask and jumps out at feature creep Jan 05 20:18:29 hey all, if I have a bbclass, modeled on angstrom-mirrors.bbclass, but with PREMIRROR instead of MIRROR, is there any reason it wouldn't try the PREMIRROR first when fetching an http source? Jan 05 20:20:29 morning Jan 05 20:21:19 Tartarus: looking at the code I see no reason for bitbake not to try PREMIRRORS first Jan 05 20:22:29 Tartarus: unless a package explicitely set PREMIRRORS="" (like p4) Jan 05 20:22:34 cyrilRomain, I don't know why either, but zlib{,-native} and fakeroot-native both go upstream / debian mirror 1st Jan 05 20:24:18 and I can see on my local mirror they exist Jan 05 20:25:07 <_diego_> hi Jan 05 20:25:16 hi Jan 05 20:25:44 <_diego_> how can i pass -pipe to gcc when building OE? Jan 05 20:27:37 hi crofton - Just checked my python version - It's 2.5.1 Jan 05 20:30:52 hi crofton - Just checked my python version - It's 2.5.1 Jan 05 20:33:57 so does anyone know if there's something linux-ish available for samsung s3c2510-based devices? Jan 05 20:35:04 Can any body see this - I'm a bit in doubt if I'm logged in currectly Jan 05 20:37:30 ssc: you are connected just fine Jan 05 20:38:07 great - thanks Jan 05 20:46:18 03coredump2 07org.oe.dev * r9bda9815... 10/ (6 files in 3 dirs): gpsd-conf: Restart gllin on a resume event Jan 05 20:48:44 join gumstix Jan 05 20:48:49 join edev Jan 05 20:48:59 join angstrom Jan 05 20:49:04 join ossie Jan 05 20:49:10 join zipit Jan 05 20:49:21 try /join Jan 05 20:50:06 Join ossie Jan 05 20:51:41 Ăthe navy Jan 05 20:53:50 quit Jan 05 20:53:51 exit Jan 05 20:53:52 end Jan 05 20:54:27 ssc: you need to use /quit and /join Jan 05 20:59:30 heh Jan 05 20:59:40 ssc was the guy with the python error Jan 05 21:00:09 Hi - I managed to connect finally - Right :-) Jan 05 21:00:19 hey Jan 05 21:00:33 I don't understand your error Jan 05 21:01:03 you should send and email to the omap list about the kernel compile issue Jan 05 21:01:16 maybe someone broke it for that board by accident ... Jan 05 21:01:22 I'm not sure I under stand it either :-), but to make a long story short. I get the the 4th last step (or so), which is do_rootfs and this one fails Jan 05 21:01:38 can you pastebin the error message Jan 05 21:01:42 ~pastebin Jan 05 21:01:43 [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste Jan 05 21:01:56 I don't understand it, but someone else might Jan 05 21:02:03 I will give it a try - Give me 5 min Jan 05 21:06:40 hereby please find a link to the last part of the error log I get http://rafb.net/p/nZb22P38.html Jan 05 21:09:06 zecke_, can you look at http://rafb.net/p/nZb22P38.html Jan 05 21:10:31 I just double checked my python version, and 2.5.1 is the latest released one unless you want to play with 3.0a2, which is still experimental Jan 05 21:10:38 that should be fine Jan 05 21:11:01 you are running bitbake minimal-image? Jan 05 21:11:36 In general, if you have a look at line 21 in the log, I think it doesn't look like a python comand line at all Jan 05 21:11:56 what is DISTRO set to? Jan 05 21:11:57 I get the same behaviour for both minimal/base/helloworld etc Jan 05 21:12:17 -n is an option for bbimage, not for python Jan 05 21:12:36 maybe `which bbimage` is badly interpolated Jan 05 21:12:51 DISTRO is set to angstrom-2008.1 - I this to new? Jan 05 21:13:45 ssc_: what is your BBPATH set to ? Jan 05 21:14:14 is -n for bbimage? Then I guess why I don't understand much of the log - Might be it should be interpret completely different than what I expect? Jan 05 21:14:55 ssc_, this works for a lot of people, we are trying to figure out what is unique about your setup Jan 05 21:15:17 BBPATH=/home/ssc/stuff/build:/home/ssc/stuff/org.openembedded.dev Jan 05 21:15:35 you should have "/usr/bin/python /home/ssc/stuff/bitbake/bbimage -n Angstrom-minimalist-image-glibc-ipk-2008.1-test-20080105-omap2430sdp -t jffs2 -e /home/ssc/stuff/org.openembedded.dev/packages/images/minimal-image.bb" Jan 05 21:15:58 hmm Jan 05 21:16:04 hi! Jan 05 21:16:05 what is your path? Jan 05 21:16:09 ssc_: check bbimage is in your PATH (it should be with bitbake) Jan 05 21:16:12 maybe bbimage is not in your path? Jan 05 21:16:33 run which bbimage from the command line Jan 05 21:16:40 my path to all the stuff is /home/ssc/stuff Jan 05 21:16:48 http://rafb.net/p/nZb22P38.html Jan 05 21:16:50 urg Jan 05 21:16:53 [balister@localhost linux-davinci-2.6]$ which bbimage Jan 05 21:16:53 ~/oe/bitbake/bin/bbimage Jan 05 21:16:58 is what I get Jan 05 21:17:24 try "which bbimage" on the command line Jan 05 21:17:27 I know, but that was the place I started - and I haven't yet moved it, althoug my /home is running out of space - Only 20 GB left :-) Jan 05 21:18:29 OK - I guess this is the problem then. I don't get anything on "which bbimage" - Hold on - I will add bitbake to my path straight away Jan 05 21:18:50 yeah Jan 05 21:19:05 we should add a check for that I suppose Jan 05 21:19:10 ssc_: that's it Jan 05 21:19:29 Crofton: no need for a check since bitbake and bbimage _has_ to be in the PATH Jan 05 21:21:01 but if you run bitbake with a path, it does fail until it looks for bbimage Jan 05 21:21:07 OK - But then I better put it there - I will have a result in a few sec Jan 05 21:21:14 Crofton: I mean it is one of the first thing new comers learns when reading GettingStarted page, setting up the environement :) Jan 05 21:21:30 yeah, but we can also fail gracefully :) Jan 05 21:21:52 Might be, but I have now done this 5 times on different linux installs, and apperent I have missed a step this time - Sorry :-) Jan 05 21:22:40 Crofton: yes, but I'm afraid the code would be bloated and could not manage _all_ the possible environement cases: bitbake in path but not bbimage, etc.. Jan 05 21:23:00 Crofton: I understand your concern though Jan 05 21:23:48 One other soultion could be to create a small script to check, that all tools are at right places, which could be run only for verifiucation of the setup Jan 05 21:24:14 ssc_: why not Jan 05 21:24:56 ssc_: but if you are ready to start a script each time you are going to work, why not running a script that set the PATH and BBPATH instead ? ;) Jan 05 21:25:38 or make sure which bbimage actually return a non-null string :) Jan 05 21:25:45 No problem - My main point is just to help new comers like me to easily figure out what they are missing Jan 05 21:26:19 ssc_: but feel free to write such a script and add it to the GettingStarted page :) Jan 05 21:26:25 if it can help .. Jan 05 21:26:26 I I guess it's really of no concern, when one is up running :-) Jan 05 21:26:30 ssc_, do you have a omap2430sdp to test on? Jan 05 21:26:33 ssc_, yeah Jan 05 21:26:50 OE is great, butting getting going can be a bit shaky ... Jan 05 21:27:03 Not at home, but I have one board at work. Not really a 2430 sdp, but a board much similar Jan 05 21:27:41 I've got a board based on a davinci processor here Jan 05 21:27:47 need to do some work on it Jan 05 21:28:15 I just finished a MS degree and reorganized my workspace at home Jan 05 21:28:25 so I am way behind on fun things Jan 05 21:28:49 As you can see at my linked in profile, I currently work for TI - Please let me know in case you miss any info - Then I can see what I can do Jan 05 21:33:59 cool Jan 05 21:34:19 I have one of the Lyrtech small form factor SDR boards Jan 05 21:34:41 which I don't know :-) Jan 05 21:34:50 I've built some images for the DV EVM, but I can't test them, although I've heard they can boot :) Jan 05 21:35:11 Why can you then test them? Jan 05 21:35:34 Aha - Now I get your point Jan 05 21:35:53 The Lyrtech card is with the davinci chip, and therefore not the DV EVM Jan 05 21:36:30 what do you miss in order to test the image on your board? Jan 05 21:37:12 I can run Angstrom on the SFFSDR board, only NFS root atm Jan 05 21:37:20 cyrilRomain and Ctofton: BINGO - Just build my first rootfs :-) Thanks for your help Jan 05 21:37:30 because no nand falsh driver in git, although that is about to change Jan 05 21:37:35 great! Jan 05 21:37:42 http://www.lyrtech.com/DSP-development/dsp_fpga/sffsdrdevelopmentplatform.php Jan 05 21:37:46 ssc_: :) Jan 05 21:38:11 who is doing the NAND flash in git? I guess I could be to great help here as well, since one of my main focuses in daily lift is flashing :-) Jan 05 21:38:36 Dirk Behme posted a patch based on someone elses work Jan 05 21:38:52 I really wish they would merge the git repos for DV and OMAP ... Jan 05 21:39:01 ssc_: In general, TI needs to open up its Linux open source development. Jan 05 21:39:19 :) Jan 05 21:39:25 I guess they have a contract with MontaVista not to release stuff early. Jan 05 21:39:31 I totally agree, but TI is a huge organisation with many interests in many places Jan 05 21:39:42 yeah Jan 05 21:39:53 ssc_: I know. But at least we are not buying the DaVinci's because of proper Linux support from TI. Jan 05 21:39:56 I talk to a lot of TI guys who wish they would open up :) Jan 05 21:40:07 Dirk Behme - He is one of the TI FEAs as far as I remember? Jan 05 21:40:19 FEA = FAE Jan 05 21:40:22 not sure who he works for Jan 05 21:40:35 I do not think he works for TI Jan 05 21:40:37 FAE = Field application Engineer Jan 05 21:40:58 I think he used to - At lease the name is sounding known to me Jan 05 21:41:27 could be, he posts from a gmail account, but I am thinking I have seen posts from a bosch account Jan 05 21:42:24 Khem, did a lot of the omap2430sdp work also Jan 05 21:42:46 I just checked, and I can only find some bosch references - Might be I mix him up with somebody else Jan 05 21:42:51 I get mixed up because I have an OSK Jan 05 21:43:13 an OSK? Jan 05 21:43:39 OMAP Starter Kit Jan 05 21:43:47 OMAP5912 based Jan 05 21:44:30 OK - But why does this mix you up with anybody? DO people think you work for TI? Jan 05 21:45:04 no Jan 05 21:45:25 but People ask me questions about hw I do not have :) Jan 05 21:45:44 OK - Now I understand :-) Jan 05 21:45:56 it has the word OMAP in it :) Jan 05 21:46:19 and the really smart guys relate omap to davinci Jan 05 21:47:08 They are as well more or less similar seen from a SW point of view :) Jan 05 21:47:46 yeah Jan 05 21:47:53 especially on the ARM side Jan 05 21:48:22 I just had a look at the link you posted for the SFFSDR - Pretty though board - My I ask for the price? Jan 05 21:48:38 I think it is like 10K usd Jan 05 21:49:05 Yeps the ARM side is pretty similar, but I think the DPS side as well isn't that different, although it's pretty hard to find some open TI DSP source Jan 05 21:49:07 a bargain! ;) Jan 05 21:49:14 hhe Jan 05 21:49:32 they claim that is close to what it costs them and cad versus usd is hurting them atm Jan 05 21:50:00 I like them because I basically got one free, and Louis bought us dinner and beer :) Jan 05 21:50:22 How did on manage to do that? By adding support for them in OE? Jan 05 21:50:51 SDR Forum radio challenge contest Jan 05 21:51:21 they haven't asked for it back yet :) I need to spend some time improving Linux support Jan 05 21:51:28 OK - Meaning you deserved it somehow - Great job :-) Jan 05 21:51:39 hopefully, I get Linux on it and get their customers to give me moeny for appliactions :) Jan 05 21:52:27 they mainly work with green hills, but they get asked about Linux occasionally Jan 05 21:52:57 my queue of things to do is pretty long, and people waving money get to the head though Jan 05 21:53:09 and at the moment, that is some other guys Jan 05 21:53:50 A what - I'm unfortunately not sure I 100% follow you Jan 05 21:55:45 I have to do some paying work :) Jan 05 21:56:48 OK - After all: money is a big thing, although it's normally not very compatible with kernel hacking :-) Jan 05 21:57:07 heh Jan 05 21:57:24 well, we need to be careful about giving away to much work Jan 05 21:57:26 Thanks for your help - Please don't hesitate to let me know in case you need any kind of favors from me, then I will see what I can do Jan 05 21:57:28 ssc_: well, look at the top kernel hackers - they get paid for it - mostly. Jan 05 21:57:52 OK - Like top Golf players :-) Jan 05 21:57:59 I work on the SFFSDR in hopes of getting pait to build on what I have done so far Jan 05 21:58:05 and it is a neat board for SDR Jan 05 21:58:16 we other guys have to pay for it and spend our limited spare time :-) Jan 05 21:59:37 the above was a continuation of my previous sentence and not a comment you your last sentence :-) Jan 05 22:00:39 I case I can do some work for OE, when I get a bit more used to it, please let me know Jan 05 22:01:28 one last question for now - I earlier today tried to build ffmpeg, which is located in OE as two different bb files Jan 05 22:02:01 one for v 0.4.9-pre1 and one for svn - It seems to me that I always try top build the old 0.4.9, which currently can build Jan 05 22:02:06 hang on, my wife just came home ,... Jan 05 22:02:14 How to I tell OE to build the svn version instead? Jan 05 22:04:28 ssc_: use PREFERRED_VERSION_ffmpeg = "svn" Jan 05 22:05:23 in local.conf Jan 05 22:05:54 Great - Thanks Jan 05 22:06:51 look in conf/distro/include for the srvrev files Jan 05 22:08:45 I will do - Thanks Jan 05 22:09:47 ok, I need to build a kernel for the sffsdr board with the nand patch and seee what happpnes Jan 05 22:10:39 I wish you great luck - In case of any nand problems you might have in the future - Please don't hesitate to ask Jan 05 22:10:49 CU Jan 05 22:12:28 my qemu-native complains about me having gcc4, not gcc3. This has never been a problem for me. (for x86 target). Jan 05 22:13:18 I'm seeing the same - What platform are you building for? Jan 05 22:44:48 * RP hates circular dependencies Jan 05 23:08:33 03Paul 07org.oe.angstrom-2007.12-stable * r29009b2f... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom-2007.1.conf: Include preferred-opie-versions-1.2.3.inc to get OPIE support. Jan 05 23:08:39 03pfalcon 07org.oe.angstrom-2007.12-stable * r0ae7b30d... 10/ (1 BACKPORTS.txt): BACKPORTS.txt: Record "angstrom-2007.1.conf: Include preferred-opie-versions-1.2.3.inc to get OPIE support." Jan 05 23:09:17 woot :D Jan 05 23:42:36 03pfalcon 07org.oe.angstrom-2007.12-stable * ra8f4ecaf... 10/ (7 files in 4 dirs): (log message trimmed) Jan 05 23:42:36 applied changes from b1eb3466dafc9eb0908b3c6a5749e44c83d43bb7 Jan 05 23:42:36 through cbd73e4685a6b456be3670606d5020189741092e Jan 05 23:42:36 gnome-mplayer: Add a decent, clean GTK+ frontend for MPlayer. Jan 05 23:42:36 * This one is targetted at simple, clean, yet functional UI and Jan 05 23:42:36 low deps (recently deps to libgnomeui and crap was removed). Obviously, Jan 05 23:42:41 its UI doesn't contain any PDA-unfriendly elements. Jan 05 23:42:45 03pfalcon 07org.oe.angstrom-2007.12-stable * r33e35a6a... 10/ (3 files in 2 dirs): Jan 05 23:42:45 Merged from .dev: Jan 05 23:42:46 b342639828f6049e03634372396237442063157a gnome-mplayer: Of course, we need MPlayer itself. Jan 05 23:42:48 153dd9902a8fbbcd4a4140f991a8e938f1af29f5 gnome-mplayer: Fix segfaults decoding non-utf8 tags in media. Jan 05 23:42:51 03pfalcon 07org.oe.angstrom-2007.12-stable * r9bdbb649... 10/ (1 contrib/angstrom/build-feeds.sh): build-feeds.sh: Add gnome-mplayer per RFC. Jan 05 23:42:58 03pfalcon 07org.oe.angstrom-2007.12-stable * rd41653e2... 10/ (1 BACKPORTS.txt): BACKPORTS.txt: Record gnome-mplayer related stuff. Jan 05 23:44:34 * flo_lap shoots blob Jan 05 23:52:48 03rpurdie * r987 10/ (4 files in 4 dirs): runqueue.py: Disable task number acceleration since it can allow the tasks to run out of sequence Jan 05 23:59:02 03rpurdie * r988 10/ (4 files in 4 dirs): runqueue.py: Improve code comments Jan 06 00:05:15 03rpurdie * r989 10/ (4 files in 4 dirs): runqueue.py: Add task scheduler abstraction and some example schedulers. Improve circular dependency chain debugging code and user feedback. Jan 06 00:54:51 03rpurdie * r990 10/ (4 files in 4 dirs): runqueue.py: Don't give a stacktrace for invalid tasks, have a user friendly message (#3431) Jan 06 00:56:15 03florian 07org.oe.dev * r5bfb1624... 10/ (5 files in 4 dirs): mainstone-kernel: Major rework of bb, activate Dyntick and tmpfs, add some modules in kernel configuration. Jan 06 01:12:21 Is it possible to name specify CONFIGFILES using a wildcard in bitbake recipes? Jan 06 01:35:38 hillct: what do you mean by CONFIGFILES ? Jan 06 01:46:08 cyrilRomain: within a bitbake recipe Jan 06 01:46:34 configuration files are specified using the CONFIGFILES directive Jan 06 01:46:56 can I specify all config files uwing a wildcard in one CONFIGFILES directive? Jan 06 01:47:05 or are wildcards not supported there? Jan 06 01:47:21 hillct: CONFIG_FILES used in autotools ? Jan 06 01:47:41 cyrilRomain: in bitbake recipes Jan 06 01:48:02 hillct: hillct I see no bitbake recipe with such a variable Jan 06 01:48:15 what? Jan 06 01:48:20 you've got to be kidding Jan 06 01:48:25 hang on Jan 06 01:48:28 lemme find one... Jan 06 01:48:55 or I'm too tired to see them (near 3am here) Jan 06 01:56:30 interesting Jan 06 01:56:53 the CONFIGFILES directive seems to have dissappeared from the reference guide Jan 06 01:57:04 perhaps it was removed in a recent version Jan 06 01:58:34 well, it's actually CONFFILES_${PN} which is why I couldn't find it... Jan 06 01:58:45 ok Jan 06 01:59:09 indeed there are plenty of them Jan 06 01:59:15 yup Jan 06 02:00:02 so, back to my original question: can I use wildcards with them or do I have to explicitly name each config file? Jan 06 02:00:08 so this is a good question Jan 06 02:00:22 * cyrilRomain is looking into bbclass files Jan 06 02:00:56 none of the examples seem to use wildcards Jan 06 02:01:03 indeed Jan 06 02:01:16 but since FILES_ use them iirc .. Jan 06 02:01:51 yeah, you'd think... Jan 06 02:02:51 hillct: at doesn't seems to support wilcards :( Jan 06 02:03:45 hillct: hmm not sure Jan 06 02:04:11 strange. Oh well. Thanks Jan 06 02:04:16 in fact the string is split and pieces are written as is in a 'conffiles' file Jan 06 02:05:00 that's classic. It accepts a list, but not a wildcard processed list... Jan 06 02:08:42 conffiles are actually used by ipk and dev packages Jan 06 02:09:17 from http://www.debian.org/doc/FAQ/ch-pkg_basics.en.html it is a list of file, and there is no mention of wildcard Jan 06 02:09:52 hillct: the final conffiles does not seems to support wildcard Jan 06 02:10:16 hillct: and bitbake does not support then also Jan 06 02:10:49 hillct: but one might create a little patch so that CONFFILES would support wildcard Jan 06 02:11:39 but that patch would involve bitbake to search the file in order to create the conffiles list (since the latter do not seem to support wildcard) Jan 06 02:17:46 hillct: so to answer your question: wildcard are not supported in CONFFILES, but the package_ipk.bbclass and package_deb.bbclass could be patched to support wildcard in CONFFILES (see what package.bbclass do with FILES_) Jan 06 02:20:51 rather more of an undertaking than I was hoping for, but probably worth doing Jan 06 02:20:55 thanks Jan 06 02:21:19 hillct: np **** ENDING LOGGING AT Sun Jan 06 02:59:57 2008