**** BEGIN LOGGING AT Sun Mar 20 23:59:56 2005 Mar 21 00:00:39 i can create one with putty now... and use that key at home also.... Mar 21 00:06:20 done Mar 21 00:10:58 is Brian Zhou here, by any chance? i have some news for him about his packages Mar 21 00:11:15 tried apache also, but that didnt work :( (was missing httpd.conf in files-dir) Mar 21 00:11:16 jp30: he was "eno" before Mar 21 00:11:23 jbot, seen eno? Mar 21 00:11:25 eno <~eno@adsl-64-164-0-84.dsl.snfc21.pacbell.net> was last seen on IRC in channel #nslu2-linux, 9h 45m 15s ago, saying: 'cool, thx'. Mar 21 00:16:58 03jp30 * 10unslung/make/w3cam.mk: fix makefile typos and added description for w3cam Mar 21 00:17:56 jp30, sorry; I was in a bit of a hurry when wedging that in. Mar 21 00:18:22 (and secretly hoping someone else would fix it. ;-) ) Mar 21 00:19:15 03jp30 * 10unslung/Makefile: promote w3cam, lua; demote units Mar 21 00:20:22 dyoung, no probs, it was quicker for me to fix it than to bug you Mar 21 00:37:09 03dyoung 07 * r1.3127 10openembedded/packages/linux/unslung-kernel-2.3r25/defconfig: Mar 21 00:37:09 defconfig: Mar 21 00:37:09 Enable video_dev=m and bring config up to post-config standard. Mar 21 00:37:14 03rwhitby * 10unslung/.cvsignore: Added tmp. Mar 21 00:38:00 Lots of action today. Mar 21 00:38:23 spring is in the air :) Mar 21 00:38:35 not here it isn't .... Mar 21 00:38:44 (autumn) Mar 21 00:38:59 :) Mar 21 00:40:59 'night all Mar 21 00:41:07 night jp30. Mar 21 00:50:47 03dyoung 07 * r1.3127.1.1 10openembedded/packages/linux/unslung-kernel-2.3r25/defconfig: Oops forgot to enable the camera device Mar 21 00:51:30 bob_tm-away, if you read the logs, that should be all the stuff in your Howto/AddUsbWebcam wiki page. Mar 21 00:52:42 so in theory if someoone updates their feed; they ought to be able to ipkg install kernel-module-videodev kernel-module-ov511 w3cam with the same net result (after rwhitby or someone updates the feed with the appropriate files) Mar 21 00:59:11 (mind you, I havnt tested any of it since I dont yet have a camera with an appropriate driver) Mar 21 01:04:03 03dyoung 07 * r1.3130 10openembedded/files/device_table-unslung.txt: Mar 21 01:04:03 device_table-unslung.txt: Mar 21 01:04:03 Add /dev/video0 Mar 21 01:08:35 sihg Mar 21 01:08:48 Can anyone else build unslung-image ? Mar 21 01:10:01 I'm getting a automake version mismatch on ipkg-native Mar 21 01:15:22 dyoung-zzzz: they will have to update to Unslung 4.5 before they can get that kernel modul Mar 21 01:15:26 e installed Mar 21 01:16:03 I'm just about to start a clean unslung build on nudi Mar 21 01:16:19 Expect a problem with ipkg. Mar 21 01:16:41 it built fine for me at home yesterday around this time Mar 21 01:19:25 03rwhitby * 10unslung/make/muxsshssl.mk: Added muxsshssl Mar 21 01:25:53 So has anyone else built 4.5-alpha yet? Mar 21 01:26:14 I would if I could, but i cant. Mar 21 01:27:35 trying to reproduce that here now ... Mar 21 01:30:35 dyoung-not-so-zzzz: ipkg-native is building now ... Mar 21 01:30:56 do_compile ... Mar 21 01:31:25 (I'm using automake (GNU automake) 1.9.4) Mar 21 01:31:39 built fine Mar 21 01:32:00 problem is at your end, sorry :-( Mar 21 01:32:47 it croaks on do_configure saying that the m4 file is version 1.8.3, but using 1.9.4 Mar 21 01:34:02 sounds like someone running automake, but not aclocal Mar 21 01:34:05 03rwhitby 07 * r1.3131 10openembedded/conf/distro/ (unslung.conf openslug.conf): Added DISTRO_TYPE to Unslung and OpenSlug Mar 21 01:34:10 I'm using GNU m4 1.4.2 Mar 21 01:34:24 rwhitby: its referring to the _m4_ macros, not m4 itself. Mar 21 01:34:36 the automake macros in aclocal.m4 are from a different version than the automake being run Mar 21 01:34:45 which indicates that aclocal wasnt run, but automake was Mar 21 01:34:46 or similar Mar 21 01:34:55 nod Mar 21 01:35:03 dunno what would be causing that here, but thats whats happening Mar 21 01:35:05 * kergoth shrugs Mar 21 01:35:07 kergoth, thats pretty much exactly what it says. Mar 21 01:35:34 dyoung-zzzz: debian? Mar 21 01:35:41 * dyoung-zzzz hides Mar 21 01:35:50 I'm using Mandrake. Mar 21 01:36:03 ooooh - new ground ... Mar 21 01:36:12 (for nslu2-linux at least) Mar 21 01:36:36 kergoth: is DISTRO_VERSION supposed to be set in the distro.conf file? Mar 21 01:37:02 ? Mar 21 01:37:08 and if so, do you bump the PR on base-files every time you change it, or do you rely on the users rebuilding base-files manually to pick it up? Mar 21 01:37:36 there is this DISTRO_VERSION variable which is used in base-files to generate the contents of /etc/issue and /etc/issue.net Mar 21 01:37:40 yes, there is Mar 21 01:38:05 i dont believe theres a set policy on whether its set by the builder or by a distro version specific .conf. Mar 21 01:38:06 is that normally set in .conf (unslung.conf or openslug.conf) ? Mar 21 01:38:08 ask one of the other distro maintainers. Mar 21 01:38:26 or use grep. Mar 21 01:38:48 kergoth, is there a way to set a variable to force-use automake version 1.8.3 (or whatever) for a specific package? Mar 21 01:38:56 dyoung-zzzz: oe? Mar 21 01:39:02 Yes. Mar 21 01:39:05 oe builds its own autoconf and automake and uses them. Mar 21 01:39:12 and does call aclocal before automake. Mar 21 01:39:17 like automake_preferred_version_for_ipkg = 1.8.x Mar 21 01:39:21 thats how it works. it never uses your local tools Mar 21 01:39:22 at all Mar 21 01:39:29 no, there isnt Mar 21 01:39:33 oh. Drat. hehe it was telling me to re-run aclocal. Mar 21 01:39:34 and it wouldnt really maek much sense. Mar 21 01:39:52 then something very strange is going on. oe calls autoreconf Mar 21 01:39:54 okaoy maybe its time to jusut blow some stuff away and see what happens Mar 21 01:40:08 autoreconf calls aclocal, autoheader, autoconf, automake, libtoolize Mar 21 01:40:13 configure.ac:3: comes from Automake 1.8.4. You should recreate Mar 21 01:40:14 configure.ac:3: aclocal.m4 with aclocal and run automake again. Mar 21 01:40:20 yes, i know what the error is. Mar 21 01:40:30 once again, that indicates that aclocal.m4 is old Mar 21 01:40:34 which makes no sense Mar 21 01:40:38 since autoreconf calls aclocal Mar 21 01:40:40 and aclocal updates it Mar 21 01:40:43 right./ Mar 21 01:40:44 Okay Mar 21 01:40:50 i've never seen that before, and have no idea what would cause that degree of a fuckup Mar 21 01:40:51 Blow away it is. Mar 21 01:40:54 i'd need to be able to reproduce it Mar 21 01:40:56 oh Mar 21 01:40:58 nevermind Mar 21 01:41:02 i do know what causes it Mar 21 01:41:02 and why Mar 21 01:41:06 i'd just forgotten Mar 21 01:41:16 * dyoung-zzzz listens Mar 21 01:41:17 dyoung-zzzz: the tmp/staging/blah/share/aclocal symlink is wrong Mar 21 01:41:27 dyoung-zzzz: the automake build doesnt overwrite it if it already exists Mar 21 01:41:30 you had a previous 1.8 build Mar 21 01:41:33 and aclocal is pointing there Mar 21 01:41:35 Aha. Mar 21 01:41:38 change it to point to 1.9 Mar 21 01:41:45 Okay I'll fix that. Mar 21 01:41:45 phil was going to fix it by getting rid of the symlink completely Mar 21 01:41:50 and the -I's pointing to it Mar 21 01:41:59 let aclocal find them in the correct version specific path Mar 21 01:42:03 which is cleaner Mar 21 01:42:09 * kergoth kicks his memory Mar 21 01:42:15 completely forgot about that :) Mar 21 01:42:25 kergoth: ok, I had already checked, and most distros do set DISTRO_VERSION in the .conf file. I was really asking whether that was the policy intention (it seems that it is). The question really is, if a variable in .conf changes, is there any way to make another package (like base-files) rebuild without the user having to manually rebuild it. I assume there isn't, but just want Mar 21 01:42:26 to check to make sure I'm not doing something wrong. Mar 21 01:43:36 no, there isnt. .conf is configuration metadata. that can change at any time Mar 21 01:43:43 the .bb doesnt track modifications to external variables coming in Mar 21 01:43:48 theres no easy way to do that Mar 21 01:44:03 ok, that's what I figured. thanks for the confirmation. Mar 21 01:44:38 i had a proto implementation in the old core that tracked md5sums of specific variables that were associated with tasks Mar 21 01:44:54 so i.e. changing SRC_URI resulted in a clean all the way up to the fetch, a refetch, then re execution back down Mar 21 01:45:00 which could be worth looking into again Mar 21 01:45:13 BTW, I did an ugly hack in unslung to allow it to install kernel-modules from OE. I added null shell scripts for /usr/sbin/update-modules and /sbin/depmod :-) Mar 21 01:45:14 this would be just another such variable Mar 21 01:45:21 ah, cool Mar 21 01:45:51 so we now have an unslung oe feed, which currently just has kernel modules in it. Mar 21 01:46:30 cool. i need to find some time and revisit that stuff. i -still- have pending target path changes that havent been pushed Mar 21 01:46:34 heh Mar 21 01:53:22 Okay that does it. Mar 21 01:53:28 Doing a full blowaway build. Mar 21 01:56:55 kergoth: any progress on uclibc 0.9.19 ? Mar 21 01:57:15 untouched, been doing a very good job at being completely and utterly unproductive this weekend Mar 21 01:58:27 now, what did I say I was going to do tonight .... Mar 21 02:00:43 I think you said you were gonna test out ctrlproxy and come up with a good sample config file. Mar 21 02:01:14 heh, no it wasn't that ... :-) Mar 21 02:03:36 guess I'll make a start on the unsling additions ... **** ENDING LOGGING AT Mon Mar 21 02:04:43 2005 **** BEGIN LOGGING AT Mon Mar 21 02:06:06 2005 Mar 21 02:06:30 dyoung: how's this look for the new unsling: Mar 21 02:06:36 Usage: $0 [flash|hdd|flash-data|hdd-data|jffs2-flash|jffs2-hdd|jffs2-flash-data|jffs2-hdd-data] Mar 21 02:06:51 defaults to external rootfs, unless you ask for jffs2 **** ENDING LOGGING AT Mon Mar 21 02:07:06 2005 **** BEGIN LOGGING AT Mon Mar 21 02:08:04 2005 Mar 21 02:08:30 rootfs in jffs2, packages on flash Mar 21 02:08:33 etc Mar 21 02:09:14 today's options now have jffs2- prepended, and the new default options put both rootfs and packages on the external disk **** ENDING LOGGING AT Mon Mar 21 02:10:06 2005 **** BEGIN LOGGING AT Mon Mar 21 02:12:04 2005 Mar 21 02:14:25 I also think that unslung 4.x should unsling to an external data partition by default too. Mar 21 02:14:41 conf is too small for the number of large packages we now have **** ENDING LOGGING AT Mon Mar 21 02:16:21 2005 **** BEGIN LOGGING AT Mon Mar 21 02:18:03 2005 **** ENDING LOGGING AT Mon Mar 21 02:19:09 2005 **** BEGIN LOGGING AT Mon Mar 21 02:19:16 2005 Mar 21 02:19:25 dyoung: comments on change of default behaviour? others? Mar 21 02:20:39 I dunno, but nslu2-log is going crazy Mar 21 02:21:05 That default behaviour is what seemed natural to me back from unslung-1.x Mar 21 02:21:21 I never really rallied for it since there were only fe packages then Mar 21 02:21:26 now I think its almost a requirement. Mar 21 02:24:12 so for unslung, the following should be all that is required, right? Mar 21 02:24:32 "/usr/bin/find / -print0 -mount | /usr/bin/cpio -p -0 -d -m -u $targ" Mar 21 02:25:01 then remove all existing /.sdXNroot files, and create the appropriate new one Mar 21 02:25:12 is that copied from the switchbox linuxrc? Mar 21 02:25:13 (only in the copy rootfs case) Mar 21 02:26:06 no, but it's the same :-) Mar 21 02:26:19 (I have it memorised :-) Mar 21 02:26:42 shouldn't need to copy /dev or /var, cause they are automatically created on boot Mar 21 02:28:37 we might want to remove the state directories from $targ too. Mar 21 02:28:54 or will does the -mount prevent them from even being copied? Mar 21 02:31:46 seemed to copy them ok Mar 21 02:34:27 one of these days someone should write openslug-switchbox-payload Mar 21 02:35:53 ah - I will need to move /dev.state to /dev etc on the target disk after I copy :-) Mar 21 02:49:09 "/dev/sdb1 19039516 45092 18800992 0% / Mar 21 02:49:09 "/dev/sdb1 19039516 45092 18800992 0% /dev.state Mar 21 02:49:09 "/dev/sdb1 19039516 45092 18800992 0% /var.state Mar 21 02:56:17 wooo hoo? Mar 21 02:56:36 I just ate my icecream, now its bedtime. Early for a change. ;-) Mar 21 02:56:54 seems to work ... Mar 21 02:57:14 I think I will remove the default options, and make people give an argument to the unsling script. Mar 21 03:08:10 03rwhitby 07 * r1.3131.1.1 10openembedded/packages/nslu2-binary-only/unslung-rootfs-2.3r25/unsling: Updated unsling script to support external rootfs in Unslung distro Mar 21 03:09:42 ok, you should be able to build 4.5-alpha, and unsling rootfs to an external disk now Mar 21 03:10:38 I've tested it enough to show that it brings up the web interface. That's the extent of my testing :-) Mar 21 03:14:47 03rwhitby 07 * r1.3135 10openembedded/packages/meta/unslung-packages.bb: Fixed unslung-packages Mar 21 03:15:55 and dyoung's camera modules are now in the unslung oe feed, and can be ipkg installed for unslung 4.5-alpha Mar 21 03:19:50 03rwhitby 07 * r1.3136 10openembedded/packages/nslu2-binary-only/unslung-rootfs_2.3r25.bb: Unslung 4.6-alpha Mar 21 03:30:06 OK, that's everything that I want to add to 4.x - you guys will have to push for anything else you want added before it goes beta .... Mar 21 03:30:39 let's aim for a feature freeze at the end of the week. Mar 21 03:55:02 hi [g2] Mar 21 03:55:43 <[g2]> hey perlguru-work Mar 21 03:56:04 i am truely openslugged now ;) Mar 21 03:56:18 bleeding edge, now Mar 21 03:56:22 <[g2]> excellent Mar 21 03:56:43 <[g2]> I'd like to try and dull that knife some :) Mar 21 03:57:07 <[g2]> less blood that way :) Mar 21 03:57:20 installed cvs also Mar 21 03:57:42 compiled samba (not tested yet) Mar 21 03:58:04 <[g2]> samba oops'ed for charta the other day Mar 21 03:58:11 <[g2]> I think it posted on the wiki Mar 21 03:59:36 where? Mar 21 04:08:42 <[g2]> http://www.nslu2-linux.org/wiki/OpenSlug/KernelPanic Mar 21 04:11:37 well, my work here is done Mar 21 04:11:39 night all Mar 21 04:21:22 <[g2]> sweet dreams rwhitby-asleep Mar 21 05:29:40 Are the instructions at this site enough to test boot openslug? - http://www.nslu2-linux.org/wiki/HowTo/TestAnImageInRamUsingRedBoot Mar 21 05:30:17 I'm using "build\tmp\staging\armeb-linux\lib\nslu2-binaries\ramdisk.ext2.gz" Mar 21 05:30:40 and "build\tmp\deploy\images\zImage-openslug". Mar 21 05:31:48 (as ramdisk.gz and vmlinuz) Mar 21 05:32:20 All I get is a beep and a red status light when I call exec. Mar 21 06:15:18 <[g2]> j0n, you'be built openslug ? Mar 21 06:15:42 <[g2]> You can just flash it, but be aware of the issues noted here http://www.nslu2-linux.org/wiki/OpenSlug/BetaRelease Mar 21 07:13:39 03bzhou * 10unslung/make/units.mk: added make/units.mk for gnu units Mar 21 07:24:48 rwhitby-dreaming: I have a suggestion - noone should be given CVS access without being registered in slugbug. Having noone to assign bugs to is frustrating. Mar 21 07:30:36 [g2], thanks. I've had nothing but red lights trying to flash the image - but I'll keep trying. Mar 21 07:36:53 bob_tm-away, agreed Mar 21 07:37:32 ka6sox-zzzz: Thanks - support is good :-) Mar 21 07:38:47 03ingeba * 10unslung/make/lame.mk: Added staging of ncurses to make lame compile Mar 21 07:38:49 bob_tm-away, np **** BEGIN LOGGING AT Mon Mar 21 07:51:03 2005 Mar 21 08:00:01 03ingeba * 10unslung/make/busybox.mk: Removed broken df applet. The one in Linksys BB works Mar 21 08:33:47 <[g2]> j0n, whan' Mar 21 08:33:57 <[g2]> j0n, when's the last time you pulled ? Mar 21 08:35:03 <[g2]> I pulled this morning and I'm flashing now. I haven't done a full build for several days though Mar 21 08:39:52 <[g2]> actually I've got an error with base-files :) Mar 21 09:16:19 [g2], last pulled about 3 days ago. Also tried about 10 days ago with similar results. Mar 21 09:16:41 I don't seem to have any problems during the build. Mar 21 10:05:32 ka6sox-away / ka6sox-office : Made http://www.nslu2-linux.org/wiki/Unslung/AddAPackageToUnslung - see slugbug references. Details in http://www.nslu2-linux.org/wiki/Unslung/PackagingBestPractices under "package registration" Mar 21 10:38:59 bob_tm-away, thanks I'll look there. Mar 21 10:39:37 ka6sox-office: Feel free to substitute your email address :-) Mar 21 10:54:46 bob_tm-away, gulp Mar 21 10:55:04 ka6sox-office: Feel free not to. Mar 21 10:55:15 <[g2]> ka6sox-office, ping Mar 21 10:55:18 bob_tm-away, np..I'm looking now Mar 21 10:55:33 [g2], hi there, how was your time. Mar 21 10:55:54 <[g2]> good. had fun thx Mar 21 10:56:00 <[g2]> how have you been doing ? Mar 21 10:56:08 excellent...I've been swamped...running ragged Mar 21 10:56:24 and we were sick for about 2 weeks it seemed Mar 21 10:56:34 <[g2]> bummer Mar 21 10:56:42 <[g2]> hope you are feeling better Mar 21 10:57:13 thanks Mar 21 10:57:28 I'm stressing at getting some work out the door by Fedex time today Mar 21 10:57:36 <[g2]> see you Mar 21 10:57:43 <[g2]> we'll talk later Mar 21 10:57:43 my assembler is here today. Mar 21 10:57:46 okay Mar 21 10:57:51 <[g2]> gl with the shipment Mar 21 10:58:03 Building latest openslug, I get a TARGET_OS set to "INVALID" and TARGET_FPU set to "". What should these be? Mar 21 13:31:27 [g2]: any plans for an OpenSlug beta this month? I'm going to push ahead on the Unslung 4.x beta now, cause it seems you've been distracted by the other development board you have ;-) Mar 21 13:32:20 I do have a later email into RR at intel Mar 21 13:32:39 outlining a plan for approval Mar 21 13:32:47 <[g2]> rwhitby-away the licensing is the main issue Mar 21 13:33:13 <[g2]> I think things look hopeful there but I wouldn't say it's all resolved Mar 21 13:34:39 cool - just letting you know that I'll be driving the 4.x beta push this month, so you guys will have to drive the OpenSlug beta in parallel. Mar 21 13:34:59 <[g2]> that's cool Mar 21 13:35:03 ok, I just forwarded the latest email communication to you two Mar 21 13:35:33 <[g2]> I'm trying to get some package stuff all straightened out Mar 21 13:35:54 <[g2]> I want to have the packages for a full development environement Mar 21 13:36:29 <[g2]> can run the tests for glibc, perl, python, ltp all on the OpenSlug tool chain Mar 21 13:36:37 <[g2]> natively on the slug Mar 21 13:37:02 ok, is that going to be a gating factor for the beta release? Mar 21 13:37:22 or if the licensing is solved the beta will go ahead immediately ... ? Mar 21 13:37:44 [g2], where do I put fertilizer? Mar 21 13:37:58 <[g2]> licensing is the #1 issue Mar 21 13:38:02 (fertilizer being the doco for OpenSlug) Mar 21 13:38:03 (native development didn't arrive until Unslung 3.x ...) Mar 21 13:38:23 <[g2]> there's an upgrade issue that's not resolved now Mar 21 13:38:38 <[g2]> except for flash/hd users Mar 21 13:38:40 03jp30 * 10unslung/Makefile: promote units Mar 21 13:38:55 <[g2]> I'm ok with the Beta users being stuck with that for now Mar 21 13:39:01 <[g2]> but it needs some more testing Mar 21 13:39:07 <[g2]> and some kinks worked out Mar 21 13:39:09 ByronT-Away: re the email, I think that we just want to be able to click through to a download and not bother with passwords. That's all intel do for the original code, so we should just do the same. Mar 21 13:39:35 <[g2]> along with stuff like settting up swap and disabling atime which is currently writing to flash sticks Mar 21 13:39:51 ok, back later ... Mar 21 13:39:58 <[g2]> nod. Mar 21 13:40:01 * rwhitby-away whistles hi ho, hi ho, .... Mar 21 13:41:04 if RR gives us the go-ahead (which I think my email has been forwarded by him to others for approval) then I have no problem with that implementation Mar 21 13:41:38 I just didn't know if we could setup a click through download at sf.net Mar 21 16:39:19 rwhitby-away: Been trying to compile the 4.x firmware this evening. Getting internal compiler error (seg fault) when compiling /home/slug/build/tmp/work/glibc-2.2.5-r4/build-armeb-linux/iconvda. Is this a known problem Mar 21 16:39:58 Sorry - lost last part of the file name: iconvdata/ibm939.os Mar 21 17:54:31 03bzhou * 10unslung/ (make/ctags.mk sources/ctags/configure.patch): added exuberant ctags for testing Mar 21 20:44:10 03bzhou * 10unslung/Makefile: ctags ready for testing Mar 21 20:47:42 03bzhou * 10unslung/ (Makefile make/bitchx.mk sources/bitchx/configure.patch): BitchX ready for testing Mar 21 23:03:02 03bzhou * 10unslung/ (make/snownews.mk sources/snownews/configure.patch Makefile): added snownews for testing Mar 21 23:45:42 morning Mar 21 23:47:31 hi there. Mar 21 23:48:24 perlguru-work, can you tell me the status of Slugbugs #18 and #80 Mar 21 23:50:36 03ingeba * 10unslung/make/libogg.mk: Makes libogg cross compile Mar 21 23:51:37 #18 OE packages for unslung: depends, do we really want OE packages for unslung? yes -> make it work. Mar 21 23:51:58 I don't think that it was the plan Mar 21 23:52:16 ok, you can reject the bug Mar 21 23:52:35 building Cross for most packages and Native for ones that cannot be built Cross. Mar 21 23:53:28 #80 using 'perldoc' as root had issues. dont know if it is still the case.... Mar 21 23:54:32 (or was it a normal user?) Mar 21 23:54:38 can you let me know if its still the case? Mar 21 23:55:11 cant at the moment: i am openslugged and dont have perl installed yet. Mar 21 23:55:27 alright..thanks. Mar 21 23:55:33 and cannot remotelly login atm Mar 21 23:55:36 np Mar 21 23:57:47 okay I'm going to reclassify #18 as "won't fix" and give an explaination. with OpenSlug Soon(tm) it might be moot. **** ENDING LOGGING AT Mon Mar 21 23:59:56 2005