**** BEGIN LOGGING AT Tue Jul 19 23:59:56 2005 Jul 20 00:02:40 Hi all. Jul 20 00:09:38 later Jul 20 00:09:49 Who be can help me? Jul 20 00:10:18 dmx_: what's up ? Jul 20 00:11:26 after command "bitbake bootstrap-image" i get "NOTE: package bootstrap-image-1.0-r0: task do_rootfs: started \n ERROR: function do_rootfs failed" Jul 20 00:11:49 task-bootstrap built with OK. Jul 20 00:13:14 task-bootstrap: unsatisfied recommendation for scap Jul 20 00:13:14 task-bootstrap: unsatisfied recommendation for kernel-module-ipv6 Jul 20 00:13:20 etc.... Jul 20 00:13:54 I it am a lot of nights try to make... :( Jul 20 00:14:37 mmm... i think the unsatisfied recommendations are not really important. Jul 20 00:14:46 What do I do not so? All under the instruction in wiki... Jul 20 00:15:24 i'm a newbie myself, so i guess you should wait for the pros to wake up ^^ Jul 20 00:15:35 but did you update your branch ? Jul 20 00:16:16 dmx_: after the missing recommends, are there any missing packets listed? Jul 20 00:16:32 03koen 07org.oe.dev * rae86a51b... 10/packages/mozilla/minimo/host_ldflags_fix.patch: Update host-ldflags patch, courtesy Patrick Steiner. Closes bug #146 Jul 20 00:18:10 koobla: No. Inside log.do_rootfs.3527 - ERROR: Cannot satisfy the following dependencies for task-bootstrap: Jul 20 00:18:13 kernel-module-bluez kernel-module-l2cap kernel-module-rfcomm kernel-module-bnep ker Jul 20 00:18:16 nel-module-hidp kernel-module-hci-uart kernel-module-sco kernel-module-bt3c-cs Jul 20 00:23:00 thats it then Jul 20 00:23:32 koen: ping Jul 20 00:25:43 rwhitby-web: pong Jul 20 00:25:52 :) Jul 20 00:26:14 koen: did you get the lat/long I sent you by email? where is that map again (URL?) ? Jul 20 00:26:14 rwhitby-web: if you ping me in an hour or 3 I'll be holding a slug :) Jul 20 00:26:31 rwhitby-web: http://ewi546.ewi.utwente.nl/maps/OE.php Jul 20 00:26:42 and five minutes after that it will have R83 removed? Jul 20 00:26:58 something like that :) Jul 20 00:27:05 * koen ponders about fattening it too Jul 20 00:27:37 koen: you spelt my nick incorrectly on the map ... Jul 20 00:28:45 D'oh Jul 20 00:28:52 heh - so I'm the only OE developer in the southern hemisphere :-) Jul 20 00:29:22 fixed Jul 20 00:29:27 above "kernel-module-*" is missed in derectory ..../build/tmp/deploy/ipk Jul 20 00:29:35 drat Jul 20 00:29:45 after "task-bootstrap" Jul 20 00:29:50 that was the mailman, and I still have no money Jul 20 00:30:00 ok, *tomorrow* I'll have my slug Jul 20 00:31:22 morning all Jul 20 00:31:27 hey RP Jul 20 00:31:37 koen: seen this: http://linuxdevices.com/news/NS6152296875.html ? Jul 20 00:32:46 koen: could you send me the source for the map page? I'd like to put one on nslu2-linux too ... Jul 20 00:33:43 morning Jul 20 00:34:52 rwhitby-web: http://pastebin.com/316837 and http://pastebin.com/316838 Jul 20 00:35:48 koen: can you change map to width:90%;height:90% instead of 400px height? Jul 20 00:36:04 hrw|work: no, the map doesn't work if I do that Jul 20 00:36:13 no idea why..... Jul 20 00:36:14 and remind url for me Jul 20 00:36:15 ;( Jul 20 00:36:33 hrw|work: http://ewi546.ewi.utwente.nl/maps/OE.php Jul 20 00:36:45 thx Jul 20 00:37:15 koen: thx Jul 20 00:37:22 rwhitby-web: you'd have to get your own key for it to work, but http://www.google.com/apis/maps/ is pretty clear on that Jul 20 00:37:34 yep, already knew about that, thx. Jul 20 00:46:45 koen: http://www.nslu2-linux.org/developers.php Jul 20 00:48:05 rwhitby-web: nice Jul 20 00:48:20 thanks to you! Jul 20 00:50:13 rwhitby-web: google uses long,lat, so you might want to swap those in OE.coords and list ($long, $lat, $text) = $pictureinfo; Jul 20 00:50:33 that way you can use 'lat long html' in the coords file Jul 20 00:54:16 koen: done: http://www.nslu2-linux.org/developers.php Jul 20 00:54:16 morning Jul 20 00:57:20 i have created a new package. when i tried to make an image it says:install: cannot create regular file `/tmp/oetmp/work/customfiles-1.0-r25/image/etc/services': No such file or directory Jul 20 00:57:27 good morning proti Jul 20 00:57:41 hi proti Jul 20 00:57:48 i have the file services in the package directory. dont know why it cant create the file Jul 20 00:58:03 rwhitby-web: And I get reminded that I should look better before thinking 'hmmm, Dyoung is hacking with the fishes' Jul 20 01:01:19 rwhitby-web: btw, you can change the red markers to slugs with GIcon Jul 20 01:01:29 Mmmm Fish.... Jul 20 01:03:44 koen: I'll leave that up to our resident nslu2-linux web developer ... Jul 20 01:03:55 :) Jul 20 01:08:00 morning Jul 20 01:08:05 hey XorA Jul 20 01:08:31 XorA: do you know how to change the default font to be installed? Jul 20 01:08:39 koen: no Jul 20 01:08:49 koen: I assumed it was in gpe-*.bb Jul 20 01:21:16 zecke: Hi Jul 20 01:21:25 OT: If I implement a 4 Bit Counter Jul 20 01:21:49 OT: should it have increment and reset, or should it just be four flip-flops saving the state? Jul 20 01:27:43 zecke, is that in VHDL? Jul 20 01:27:50 Or a generalised question? Jul 20 01:32:34 dyoung: it is for a GAL Jul 20 01:32:52 dyoung: but I'm supposed to program a 4-bit counter Jul 20 01:33:00 dyoung: should it be able to store four bits Jul 20 01:33:10 dyoung: or should have it alu functionality as well? Jul 20 01:34:53 zecke: make sure the counter also does mmx and takes x86 asm Jul 20 01:35:20 koen: Why do you always name the obvious requirements? Jul 20 01:36:00 :) Jul 20 01:36:25 Sadly internal it is only and ARM Version10 core Jul 20 01:36:37 the multi core specification came too late to bei implemented in my GAL Jul 20 01:36:59 not arm11? Jul 20 01:37:18 or does arm10 support armv6? Jul 20 01:37:32 I fear not, maybe we get more government funding... Jul 20 01:37:58 Yeah, having a ALU would be nice. Jul 20 01:38:22 dyoung: psst, I'm using four damn RS Flip Flops Jul 20 01:38:32 we've implemented them the task before :} Jul 20 01:38:41 * zecke is so fucking lazy Jul 20 01:38:59 koen: what is KERNEL_VERSION ?= kernel24 supposed to do? Jul 20 01:38:59 FF's are you friend. Jul 20 01:39:04 koen: is it in the overrides? Jul 20 01:39:16 zecke: yes Jul 20 01:39:26 zecke: just read the whole .conf Jul 20 01:39:44 strange, I do not want to believe it ;) Jul 20 01:40:05 maybe my parser move broken more than I had expected (maybe it is silenting errors?) Jul 20 01:42:21 World.setDoubled(false) Jul 20 01:44:12 hi Jul 20 01:44:24 hey do13 Jul 20 01:44:46 hi koen Jul 20 01:45:16 zecke: Related to koen problems, are we allowed to override OVERRIDES ? Jul 20 01:45:52 proti: override OVERRIDES as in OVERRIDES_foo = "yippe" Jul 20 01:45:59 Yes. Jul 20 01:46:04 proti: or in data.setVar('OVERRIDES', 'moo') Jul 20 01:46:15 proti: technically yes, but I wonder what happens ;) Jul 20 01:46:20 No, like OVERRIDES_prepend ="toto:" Jul 20 01:46:39 _prepend is not an override Jul 20 01:46:45 It needs 2 update_data call to get the result right. Jul 20 01:47:48 if that is the case it is a bug, create a test suite please ;) Jul 20 01:48:02 probably only using the data module directly Jul 20 01:48:24 zecke: MACHINE=h3900, bitbake virtual/kernel Jul 20 01:49:10 koen: no good testcase Jul 20 01:49:21 koen: what about KERNEL_VERSION = or := instead of ?= Jul 20 01:49:27 LOL : moon.google.com Jul 20 01:49:32 zecke: update_data gets the OVERRIDES value then do the applyOverrides and then do the the _append and _prepend values. Jul 20 01:50:04 proti: well then ;) Jul 20 01:50:36 let us forbid _append _prepend for OVERRIDES Jul 20 01:50:42 who uses this anyway? Jul 20 01:50:45 a word of warning: i found ?= to be not working at all sometimes, so perhaps this shields another bug Jul 20 01:50:49 morning, btw. Jul 20 01:50:49 koen Jul 20 01:50:51 and bbl Jul 20 01:51:00 mickey|ttg500: AFAIK ?= should do nothing ;) Jul 20 01:51:11 mickey|meeting: but we've woken kergoth up by now Jul 20 01:51:31 zecke: h3600.conf and h3900.conf, and possibly others Jul 20 01:51:59 koen: lol zooming maximum is nice Jul 20 01:52:04 zecke: I was thinking of doing the override and append_prepend data to OVERRIDES first and then doing it for the rest of the array. Jul 20 01:52:12 zecke: say cheese :) Jul 20 01:52:29 proti: how to handle _append_foo Jul 20 01:52:42 proti: do you want two cases for _append and append with overrides? Jul 20 01:52:58 simply do not do it with overrides Jul 20 01:53:16 need to finish my protocol now Jul 20 01:53:59 zecke: http://www.google.com/jobs/lunar_job.html Jul 20 02:01:15 re Jul 20 02:01:54 hehe Jul 20 02:02:14 * zecke is turning the assignment in now Jul 20 02:03:39 cya Jul 20 02:15:16 hi Jul 20 02:16:14 hey Bernardo Jul 20 02:18:53 hi koen, porting maemo to a 640x480 screen? :) Jul 20 02:20:10 Bernardo: no, I was too lazy to setup Xephyr :) Jul 20 02:20:47 :) Jul 20 02:21:10 hi Bernardo Jul 20 02:21:13 I still have to get monotone working on my machine to start working on the c1000 port of oz again... Jul 20 02:21:15 hi hrw|work Jul 20 02:21:24 how were your hollidays? Jul 20 02:21:32 Bernardo: too short but nice Jul 20 02:21:44 I still have to make serial cables ;( Jul 20 02:22:35 And I still have to get monotone working... I am using the sharp rom for now, unfortunately... Jul 20 02:25:22 hi BigAl Jul 20 02:25:29 Hi hrw Jul 20 02:30:35 morning all Jul 20 02:33:10 hi ade Jul 20 02:35:19 rebuilding oz 3.5.3 again ;( Jul 20 02:54:16 i am trying to build an image with base-files package. i have created a new package which creates some of the directories like home etc proc. i am not finding all of them getting reflected in my file system. i am getting install -m 2755 -d /tmp/oetmp/work/customfiles-1.0-r25/image/home when i do bitbake -D but it does not get reflected in the file system. Jul 20 02:54:51 eJumbo: added that into FILES? Jul 20 02:54:52 sorry, image without the base-files package. i am experimenting i want my own filesystem with my own entries. Jul 20 02:56:18 hrw|work: where do i find that ?. in the .bb file ? Jul 20 02:59:20 yep Jul 20 03:00:16 re Jul 20 03:03:21 can find FILES_${PN} in base-files.bb but those directories are not been assigned to it. Jul 20 03:04:22 eJumbo: so you have an answer Jul 20 03:05:14 sorry hrw| cannot find them Jul 20 03:05:37 misspellt can Jul 20 03:11:40 what does FILES_${PN} = "/" do ?. Jul 20 03:11:56 03law 07org.oe.dev * re914b209... 10/packages/linux/linux-openzaurus-2.6.12-mm1/defconfig-c7x0: updated kernel conf for clamshell ppp should work again Jul 20 03:12:03 if i insert it. i get all my directories in the file system ?. why ? Jul 20 03:12:13 eJumbo: You seems to have an bitbake. Could you update it please ? Jul 20 03:12:50 mine is working ?. i am happy with it ? Jul 20 03:12:57 :) Jul 20 03:13:10 Where lies the FILES_${PN} = "/" ? Jul 20 03:13:21 __law__: could you explain with more details next time when pushing? Jul 20 03:13:30 base_file Jul 20 03:14:11 if i remove it. some directories are not created like proc. if i insert that line it creates the directories. why ? Jul 20 03:14:32 <__law__> hrw|work, okay Jul 20 03:15:40 __law__: for me better way would be checking why ppp does not work in modules - c7x0 kernel space is limited so sooner or later ppp will again get lost from kernel Jul 20 03:15:56 __law__: and you forgot to bump linux-openzaurus_2.6.12-mm1.bb Jul 20 03:18:14 eJumbo: Seems to set the base prefix to the installation. Jul 20 03:18:30 <__law__> hrw|work, there isnt any change on the .bb file Jul 20 03:19:19 __law__: defconfig changed so kernel changed - to force rebuild new kernel you have to bump PR Jul 20 03:19:52 __law__: for me you changed kernel so I have to be forced to rebuild it == you have to bump PR Jul 20 03:20:16 <__law__> hrw|work, ah ok Jul 20 03:20:36 hm Jul 20 03:20:36 the general rule is that if the generated ipkg(s) change, you have to bump PR Jul 20 03:20:48 ~lart _law_ Jul 20 03:20:48 * ibot slaps _law_ upside and over the head with one freakishly huge killer whale named hugh Jul 20 03:20:54 hey you can't just tweak a kernel defconfig Jul 20 03:21:04 this kernel defconfig is hand tuned Jul 20 03:21:48 a) this change is not isolated, i.e. perhaps we need to change c7x0.conf to adapt and b) there was a reason for us to ship these things as modules. the reason was: size. the kernel may be too large now again Jul 20 03:22:05 i prefer if you would discuss this with the .bb file maintainers Jul 20 03:22:08 in other way - do not push changes without contact with MAINTAINER Jul 20 03:23:20 bbl lunch Jul 20 03:23:38 <__law__> __law__ ppp doesnt work for me with the current oz kernel (2.6.12-mm1) but if i include ppp stuff direct into the kernel (no modules) all works fine Jul 20 03:23:44 <__law__> __law__ should i release my defconfig-c7x0 ? Jul 20 03:24:47 What's the reasons for ppp not working if modular? Jul 20 03:25:24 who did you talk with? Jul 20 03:25:44 <__law__> mickeyl, they wound load Jul 20 03:25:54 ok, any more specifics? Jul 20 03:26:03 why wouldn't they load? symbols missing? crash? Jul 20 03:26:36 proti: did not get u ? Jul 20 03:26:59 <__law__> i havent installed this kernel anymore, so i cant reproduce it yet Jul 20 03:27:04 * XorA manages to break firefox ipk even worse :-) Jul 20 03:27:15 hmm Jul 20 03:27:22 ok you see that was not the right workflow Jul 20 03:27:51 XorA: breaking stuff is easy ;) Jul 20 03:27:57 <__law__> yes should i step back to the old config version? Jul 20 03:28:05 firefox is easier than most... Jul 20 03:28:45 koen: breaking ipkg seems to be easy :-) Jul 20 03:28:48 _law_: leave it for some time until i have checked why the modules don't load Jul 20 03:29:18 mickey|lunch: smacznego Jul 20 03:30:03 <__law__> mickey|lunch, ok i?ll push the new bb file Jul 20 03:37:03 03law 07org.oe.dev * rb1e95c69... 10/packages/linux/linux-openzaurus_2.6.11.bb: updated kernel conf for clamshell ppp is now directly build into the kernel Jul 20 03:37:48 __law__: you pushed change for 2.6.12-mm1 defconfig and then bumped 2.6.11? Jul 20 03:42:08 03law 07org.oe.dev * r95b99e89... 10/packages/linux/ (linux-openzaurus_2.6.11.bb linux-openzaurus_2.6.12-mm1.bb): fixed typos Jul 20 03:57:06 proti: who is doing the OVERRIDES_prepend. this sounds like a stupid idea anyway Jul 20 03:57:19 either do := assignment or don't do it all Jul 20 03:58:55 argh.. Jul 20 03:59:24 syntax error near unexpected token `'INHIBIT_PACKAGE_STRIP',' usually means "you forgot to inherit package_ipk" - right? Jul 20 03:59:31 hi all again Jul 20 03:59:35 hrw|work: yes Jul 20 03:59:49 zecke: like I said, h3600.conf, h3900.conf and possibly others Jul 20 04:00:03 ~lart recent bitbake or something other Jul 20 04:00:04 * ibot teaches recent bitbake or something other that M$ Access is a database. No, really, a database. A real live multi-user... well, ok, not multi-user, but a database. Yeah, that sounds right. Jul 20 04:00:23 ERROR: Cannot satisfy the following dependencies for task-bootstrap: kernel-module-hidp kernel-module-bt3c-cs . Where defined this depends? Jul 20 04:02:29 koen: that is stupid Jul 20 04:02:53 zecke: talk to pb_ Jul 20 04:03:37 dmx_: task-bootstrap RDEPENDS on these packages in someway Jul 20 04:03:43 pb_: master? Jul 20 04:04:06 hrw|work: where does that come from? Jul 20 04:06:31 zecke: I do one thing which maybe is wrong - trying to build oz 3.5.3 with OE from 3.5.3 and current bitbake Jul 20 04:06:53 hrw|work: that could be the issue Jul 20 04:07:14 maybe Jul 20 04:07:28 oz 3.5.3 was before the restructuring in bitbake? Jul 20 04:07:33 yes Jul 20 04:09:02 zecke: at your service Jul 20 04:18:18 pb_: Thanks to proti we know what is going ;) Jul 20 04:18:48 pb_: OVERRIDES_prepend = "${KERNEL}:" is evaluated only at the end of update_data() Jul 20 04:19:09 pb_: is there a reason not to put KERNEL in OVERRIDES in bitbake.conf? Jul 20 04:19:23 pb_: or why did you chose _prepend instead of immediate asisgnment? Jul 20 04:20:38 zecke: no reason. I think using immediate assignment would be fine. Jul 20 04:23:39 pb_: or do you think OVERRIDES_prepend should work? Jul 20 04:30:12 zecke: "R" in RDEPENDS = REMOVE? Jul 20 04:30:20 zecke: from an idealistic point of view, I think that OVERRIDES_prepend should work. But I can't think of any good way to make it work without making update_data even slower than it already is. Jul 20 04:30:48 dmx_: my abbreviation is R for Remote Jul 20 04:31:08 Runtime DEPENDS Jul 20 04:31:20 Russian DEPENDS Jul 20 04:31:21 lol Jul 20 04:31:30 yeah, russian is correct Jul 20 04:31:36 yeah Runtime was my abbreviation as well Jul 20 04:31:54 pb_: that is my pragmatic point of view as well Jul 20 04:33:27 yeah. I'm happy to forbid the application of overrides to OVERRIDES itself, in the interests of an easy life. Jul 20 04:34:19 I think the only straightforward way to get the completely correct behaviour would be for update_data to repeatedly apply overrides until there are no further changes to the data set. Jul 20 04:34:54 But that would always require at least one extra scan over the dictionary (to determine that there were no changes), which seems like bad news. Jul 20 04:36:33 looking at OVERRIDES flags might suffice as well? Jul 20 04:36:55 well, the trouble is that you could have overrides applied to variables that are referenced by OVERRIDES, which would cause the same problem. Jul 20 04:38:03 right Jul 20 04:38:22 so you'd need to inspect the flags recursively. I guess that wouldn't be impossible. Jul 20 04:38:46 we could run GNU/Hurd from within emacs started by bitbake to do checking Jul 20 04:38:53 that would be a nice research topic Jul 20 04:39:05 * koen shrugs Jul 20 04:39:32 running the hurd inside emacs is always an option Jul 20 04:44:00 ~fishslap zecke for setting OLDEST_KERNEL to 2.6.11 Jul 20 04:44:01 * ibot slaps zecke up side the head with a wet fish for setting OLDEST_KERNEL to 2.6.11 Jul 20 04:44:16 zecke: i remove depends by hands from zaurus-clamshell-2.4.conf, removed direcotories in build/wordk, input "bitbake task-bootstrap" and "bitbake bootstrap-image". Error repeated. Jul 20 04:45:00 dmx_: maybe the bbs are still kept in the cache dir Jul 20 04:45:44 hmm.. Jul 20 04:47:33 zecke: heh Jul 20 04:49:20 which analog "make clean" or "make distclean" in bitbake? Jul 20 04:49:35 re Jul 20 04:49:49 re? Jul 20 04:50:03 what is re?... Jul 20 04:50:03 dmx_: Yes, I'm back. Jul 20 04:50:07 freyther@stabilo:~/programming/qemu/qemu-0.7.0$ ../arm/cat Jul 20 04:50:07 FATAL: kernel too old Jul 20 04:50:13 proti: :) Jul 20 04:50:44 dmx_: Did you test bitbake -c clean ? Jul 20 04:52:17 pb_, zecke: So ? Shall the OVERRIDE_{prepend|append} be forbidden ? Shall we treat the OVERRIDES var in a special way ? Jul 20 04:52:39 ok, i will be trying it... Jul 20 04:52:54 proti: Two votes for forbidding Jul 20 04:53:22 zecke: nope, 3 votes for forbidding :). Jul 20 04:53:44 and I've casted my vote for improvinf update_data Jul 20 04:54:34 That is a delicate matter. Jul 20 04:56:26 I've put a lot of thoughts about improving its performance. I'm afraid it won't be straightforward. Jul 20 04:56:27 pb_: mov r7,#798 should work? Jul 20 04:56:53 proti: Improve by not calling update_data Jul 20 05:00:06 zecke: Just parsing invokes, expandKeys, update_data, set_additionnal_vars, update_data for each bbfile. That is the main culprit of slow parsing. Jul 20 05:00:43 proti: set_additional_vars does not look heavy though Jul 20 05:00:59 zecke: Yes, but it forces an update_data after. Jul 20 05:02:01 http://pastebin.ca/18124 is result of that bug which you talk about? Jul 20 05:02:26 pb_: When is glibc 2.3.5 landing in unstable? Jul 20 05:03:01 hrw|work: possible Jul 20 05:03:06 bb all Jul 20 05:03:18 thanks a lot Jul 20 05:03:34 hrw|work: peek into modutils-cross Jul 20 05:03:39 hrw|work: and see what it provides Jul 20 05:05:06 virtual/arm-linux-depmod Jul 20 05:05:40 hrw|work: so it is a old modutils-cross and new kernel? Jul 20 05:05:52 kernel 2.4.18/collie I want Jul 20 05:05:59 anyway with poke maybe it will go Jul 20 05:06:09 ;) Jul 20 05:06:24 :) Jul 20 05:06:25 mickey|lunch: I request poke to be removed, people exploit it in ways... ;) Jul 20 05:06:32 it does not ;( Jul 20 05:07:01 zecke: let it exist but not show in help Jul 20 05:16:11 I have weird idea.. installing jabberd on wrt54 Jul 20 06:11:43 pb_: can you explain the hook in do_split_packages for Jul 20 06:15:36 OE core devs: We have busybox and coreutils which both provide a number of programs (e.g. "cat"). Jul 20 06:15:54 The trouble is that busybox puts it in /bin/cat, and coreutils puts it in /usr/bin/cat. Jul 20 06:16:22 The FHS says "cat" should go in /bin: http://www.pathname.com/fhs/pub/fhs-2.3.html#REQUIREMENTS2 Jul 20 06:16:37 looks like we need to patch coreutils Jul 20 06:16:58 We want to add update-alternatives to coreutils, so that you can install busybox and coreutils, and have the overriden files pointed to coreutils. Jul 20 06:17:20 zecke: mm? Jul 20 06:17:21 So we patch coreutils and add update-alternatives for all the conflicting files. Is that appropriate? Jul 20 06:17:36 rwhitby: yeah Jul 20 06:17:41 Similarly for grep and tar and .... Jul 20 06:18:32 Do you think anyone will be upset if programs that used to be in /usr/bin suddenly get moved to /bin instead? Are /usr/bin and /bin on different storage media for any of the OE distros, or anything like that? Jul 20 06:18:41 * koen imagines we need a lot of CONFFILES too Jul 20 06:18:47 pb_: thanks Jul 20 06:18:57 zecke: any time Jul 20 06:19:27 koen: conffiles are problem for gui packagemanagers :( Jul 20 06:19:32 rwhitby: having binaries drift around the filesystem is a bit unfortunate, but I think conforming to the FHS is a worthwhile enough justification. Jul 20 06:19:42 hrw|work: so? Jul 20 06:19:50 rwhitby: I don't think anyone is entitled to get too upset about that. Jul 20 06:19:51 rwhitby: could you also look into busybox<>console-tools alternatives too? Jul 20 06:20:10 DaKa2 is the one doing the work :-) Jul 20 06:20:11 koen: nothing - opie-pm is nearly unusable anyway Jul 20 06:20:14 hehe Jul 20 06:20:21 ~hail DaKa2 Jul 20 06:20:22 * ibot bows down to DaKa2 and chants, "I'M NOT WORTHY!!" Jul 20 06:20:28 I'll look into console-tools too Jul 20 06:21:15 OK, so our mantra is that any other package which conflicts with busybox is given a higher priority than busybox for update-alternatives so that the non-busybox package gets priority. Jul 20 06:21:58 pb_: the background to all this is that we are putting together a native development environment for OpenSlug firmware on the NSLU2. Jul 20 06:23:07 The goal being to be able to run bitbake with OE metadata on the slug itself to build some packages natively. Our preference is still cross-compilation of course, but sometimes you just have to build something native ... Jul 20 06:23:45 So we want to create a meta/task-openslug-development.bb which has lots of packages which override busybox programs. Jul 20 06:23:47 rwhitby: or atleast run configure scripts to populate config.cache (site/*) Jul 20 06:24:23 hrw|work: yes, for most packages you can do that. Some things (e.g. emacs "dump") are a bit more difficult to do cross ... Jul 20 06:25:30 (not impossible, just more difficult than most developers are happy to invest time in ...) Jul 20 07:06:49 eek update-rcd bbclass is... Jul 20 07:08:10 evil? Jul 20 07:10:21 unexpected Jul 20 07:10:42 I had a syntax error in my python construct but it complained about rcd_update* Jul 20 07:12:38 working on the c-parser? Jul 20 07:13:04 koen: no package splitting Jul 20 07:13:14 and work? Jul 20 07:13:23 what a question, I'm lazy Jul 20 07:19:26 hi Jul 20 07:19:33 hi Jul 20 07:51:12 time to go - cu Jul 20 08:10:30 Is vanille.de down? Jul 20 08:10:54 try the other hosts Jul 20 08:11:01 the wonder of monotone Jul 20 08:12:52 For fear of destroying my db: I usually do 'monotone --db=./oe.db pull'. I guess 'monotone --db=./oe.db pull ewi546.ewi.utwente.nl "org.openembedded.{dev,dreambox,nslu2-linux}" ' should work fine? Jul 20 08:13:09 yes Jul 20 08:13:17 Thanks for the confirmation. Jul 20 08:13:20 monotone is completely distributed Jul 20 08:15:41 _law_: more changes needed for your ppp changes Jul 20 08:17:19 <_law_> XorA: ? Jul 20 08:18:14 _law_: was you moved ppp into kernel? Jul 20 08:19:58 <_law_> XorA: ppp kernel stuff is compiled directly into kernel, not as module Jul 20 08:20:49 _law_: yes, but you also need to fix the zaurus-clamshell-2.6.conf to not depend on ppp as a module Jul 20 08:27:54 hey all, i have a few questions about how the initrd works on linux/arm Jul 20 08:30:44 who knows about arm and linux specifics? Jul 20 08:30:59 /* Put the right bits in RAM so the kernel will find the initrd */ Jul 20 08:30:59 #if defined(__i386__) Jul 20 08:30:59 boot->setup->ramdisk = initrd_addr; Jul 20 08:30:59 boot->setup->ramdisk_size = size; Jul 20 08:31:17 this snippet of code controls the function of Two Kernel Monte's user space program Jul 20 08:31:33 i'm wondering, what do i put for #if defined(__arm__) Jul 20 08:31:41 it only has i386 and alpha right now Jul 20 08:39:16 <_law_> XorA|gone: changed Jul 20 08:41:47 03law 07org.oe.dev * r6da8c85d... 10/conf/machine/zaurus-clamshell-2.6.conf: removed ppp-modules from openzaurus-2.6.conf. because ppp is build fix into the kernel Jul 20 08:44:02 howdy people Jul 20 08:44:36 howdeeeeee! Jul 20 08:45:04 sup? Jul 20 08:45:14 btw where were you again koen? Jul 20 08:45:48 52.23410,6.846017 :) Jul 20 08:46:03 which is where :P Jul 20 08:46:19 enschede, the netherlands Jul 20 08:52:42 ahh, is there anywhere near that "What the Hack" conference thingy? Jul 20 08:55:42 * mithro beats wireless with a stick Jul 20 08:59:13 Marcel Holtmann just said that people are working on Bluetooth via SD Jul 20 08:59:18 anyone here working on thateffort? Jul 20 08:59:33 not me Jul 20 08:59:58 maybe the OpenBSD guys Jul 20 09:00:15 http://www.nrk.no/nyheter/innenriks/4913396.html Jul 20 09:00:20 woops Jul 20 09:00:37 didn't mean to paste that here Jul 20 09:01:06 where could i find an ARM guru? Jul 20 09:01:07 hmm... I interested in seeing if anybody has gotten their hands on a card spec Jul 20 09:01:13 jnc: visit #elinux Jul 20 09:01:22 'k Jul 20 09:01:27 I also type bad real Jul 20 09:02:31 real bad Jul 20 09:02:38 ;) Jul 20 09:03:36 head to munich tommorrow Jul 20 09:03:44 Harlekin hasn't replied yet :/ Jul 20 09:06:10 CP|Laptop: that's BT via SDIO, right? Jul 20 09:06:17 koen: yeah Jul 20 09:07:06 the 'cguys' claim their product works under linux Jul 20 09:07:17 heh Jul 20 09:07:19 yeah, I'm sure Jul 20 09:08:04 I want to kick start SDIO development, but I can't really do anything because we're under NDA with the SD association Jul 20 09:08:08 binary only as such Jul 20 09:08:22 So I want to provide as much moral support as I legally can to others Jul 20 09:08:41 whoops, my laptop got stolen with all the specs on it Jul 20 09:08:44 damn.... Jul 20 09:09:11 heh Jul 20 09:09:20 koen: stolen or "borrowed" ? Jul 20 09:09:40 The thing about SDIO is that its not really interesting without insight into the cards on the other side Jul 20 09:15:10 CP|Laptop: if you provide me with a device I might work on a driver ;P Jul 20 09:15:42 Hello, could some one please tell me if there is a difference in the way serial ports are handled in familiar to a normal desktop linux dist. - as non of the ipaq serial ports show up in my test prog, which makes all the ones on my desktop visable Jul 20 09:15:50 mithro: you serious? Jul 20 09:16:44 CP|Laptop: well kinda, i have enough things to do - but if somebody was to provide me with a bunch of SDIO devices I might have to get down and write something Jul 20 09:18:09 mithro: do you have any particular devices in mind? Jul 20 09:18:39 DVB-T and DVB-H adapters ;) Jul 20 09:18:43 what are there? only three I know are SDIO wifi, SDIO bluetooth and SDIO FRM radio Jul 20 09:18:55