**** BEGIN LOGGING AT Sun May 08 23:59:57 2005 May 09 00:09:25 wasnt there a freeradius ipkg at some point? May 09 00:10:46 I'm pretty sure someone was working on one May 09 00:11:09 listed in "needs to be fixed". dunno why May 09 00:11:17 listed as working for wiley May 09 00:11:42 ok.. I can test :) May 09 00:12:14 I want to put up a bunch of slugs as radius proxies.. diskless May 09 00:13:09 what is the best way to build for unslung btw.. openembedded? May 09 00:13:40 packages? May 09 00:13:41 Optware May 09 00:13:42 yes May 09 00:13:45 optware May 09 00:13:48 hm, ok May 09 00:13:55 (nee Unslung package build system) May 09 00:15:40 anyways, if there is a freeradius ipkg somewhere, I'd like to test it right a way :) May 09 00:22:19 (hm, why is there a wl500g folder under the unslung feed?) May 09 00:23:17 cause Optware has been ported to wl500g too May 09 00:23:17 ok, found the freeradius ipkg May 09 00:24:00 hm, never heard of optware.. I need to read up May 09 00:24:13 christopher initially checked it in. May 09 00:24:38 New name for Unslung packages. Built for Unslung on NSLU2 and Oleg's firmware on WL500g (WileyWare) May 09 00:25:23 ok, cool... I have a wl-hdd too :) May 09 00:25:29 NOTE: package flex-native-2.5.31-r2: task do_populate_staging: started May 09 00:25:29 ERROR: function do_stage failed May 09 00:25:50 mf.. no room left for freeradius :) May 09 00:25:59 so flex is biting me. May 09 00:26:20 hehe, it was room for freeradius, but not for the ipkg meta data May 09 00:27:07 so now I have a inconsistant ipkg database :) May 09 00:27:11 ka6sox it's sourceforge's fault - I just downloaded it manually and put it in downloads dir May 09 00:27:43 so just get it from somewhere else? got a url where you go it from? May 09 00:27:52 then I'll do the same thing. May 09 00:29:50 I just copied it from my other downloads dir May 09 00:30:58 i cant find a working url for it May 09 00:32:54 http://easynews.dl.sourceforge.net/sourceforge/lex/flex-2.5.31.tar.bz2 May 09 00:34:27 ka6sox: this is the same as [g2]'s problem, his download was ok (check the .md5), but he had CXX and CC set in his environment May 09 00:34:45 I'm on a debian system. May 09 00:35:24 flex-2.5.31.tar.bz2?download&failedmirror=telia.dl.sourceforge.net May 09 00:35:24 makes no difference I believe - it's not gentoo, it's the environment. Something in that version has made the build sensitive to it, somehow. May 09 00:36:14 the flex download is short. May 09 00:36:23 and it fails md5 May 09 00:37:39 'morning May 09 00:39:19 ka6sox: um, what's more both do_configure and do_compile explicitly set CXX and CC. Unless it was LC_CTYPE (unlikely) I think the environment variable explanation is wrong... May 09 00:40:05 morning. May 09 00:40:51 ka6sox http://easynews.dl.sourceforge.net/sourceforge/lex/flex-2.5.31.tar.bz2 works May 09 00:40:55 for sure the telia.sf.net file was broken May 09 00:41:07 okay I've put it manually in downloads May 09 00:41:12 and restarting May 09 00:42:42 ERROR: function do_stage failed May 09 00:42:42 ERROR: see log in /home/tking/svn/openslug/openslug/tmp/work/flex-native-2.5.31-r2/temp/log.do_stage.837 May 09 00:43:16 install: cannot stat `flex': No such file or directory May 09 00:43:58 simple answer, try removing tmp dir and starting again May 09 00:44:12 the md5 has 0 length. May 09 00:44:14 less simple, remove the flex dir in tmp/work and the flex stamps May 09 00:44:33 hm, I can delete most of the rlm libs May 09 00:45:19 ka6sox: the error message sequence is identical (both cases) to that which [g2] experienced. May 09 00:45:48 So, I still say: cd / * Pick out the memory size. We look for mem=size@start, May 09 00:45:48 * where start and size are "size[KkMm]" May 09 00:46:01 Oops. Can't copy'n'paste May 09 00:46:41 okay...iirc the memory is continguous. May 09 00:47:01 I was attempting to say use jacques second approach (cd openslug/tmp, rm -rf */flex-native-2.5.31-r2) May 09 00:47:09 rm -rf tmp May 09 00:47:11 make May 09 00:47:27 ;-) May 09 00:48:18 thats almost why I want a static cache...about 1/2 the time I end up having to manually intervene and find the source since the script is fixed :P May 09 00:48:20 I believe if any step of a package fails then it's always sufficient (after fixing the problem) to kill just the temp/foo directory and the temp/foo.do_* stamps. May 09 00:48:35 my build is long past flex so I'm confident you too can get past it May 09 00:49:04 I'm still worried about the md5 that has 0 length. May 09 00:49:35 It doesn't check the md5 or it wouldn't have failed in the first place... May 09 00:49:47 yeah, I consider that a bug May 09 00:50:08 it should have failed at the download, not after trying to untar a nonexistent file May 09 00:50:46 that makes more sense and gives me better confidence that we have good sources. May 09 00:51:58 maybe oe needs additional smarts to detect a failed sf.net download, since wget does not exit abnormally in that case May 09 00:52:03 on to bison....it made it thru flex. May 09 00:52:28 look at the flex md5 file now - is it 32 bytes? May 09 00:53:17 no...still 0 May 09 00:53:34 thats a bug May 09 00:54:43 mine somehow became 32 bytes, ah well a minor mystery May 09 00:57:49 it's so cool to see bitbake taking < 100MB May 09 00:58:11 (I've never used the -symlinks stuff before) May 09 01:02:02 It's about 250k/.bb file. openslug-packages is 116 .bb files, so that's 29MByte plus overhead. There are 92 package *directories*, containing 387 .bb files, so with symlinks it takes about 97MByte - about 3 times more memory than it absolutely needs. May 09 01:04:19 jbowler, sometimes you scare me May 09 01:04:22 :-D May 09 01:07:55 Ah well, I run a script which finds all the .bb files which were used in the build, then next time I just use those. Number of lines in script output = number of .bb files... May 09 01:10:19 yeah I saw mention of your script on the list I think May 09 01:13:30 Attempting a build now. May 09 01:15:11 dyoung you will probably have same flex problem as the rest of us May 09 01:15:30 you can manually dl it from http://easynews.dl.sourceforge.net/sourceforge/lex/flex-2.5.31.tar.bz2 to your downloads dir May 09 01:20:02 hmmm, my build is having a hell of a time trying to dl fakeroot May 09 01:20:19 going through every single debian mirror May 09 01:20:33 all over the world May 09 01:20:41 might be too old May 09 01:21:06 personally, I can copy things from my many and various downloads dirs, but new builders might have problems May 09 01:21:53 ah, OK it finally got fakeroot from http://www.oesources.org/source/current/fakeroot_1.2.2.tar.gz May 09 01:22:10 That's the new OE mirror (the treke replacement). May 09 01:22:35 ah cool May 09 01:22:49 I'm glad someone is keeping abreast of #oe May 09 01:25:22 lol, it tried 23 debian mirrors looking for fakeroot before it got to oesources May 09 01:27:03 DL_DIR in local.conf.sample is ${HOME}/sources, that maybe makes much more sense than ${OEROOT}/downloads May 09 01:36:04 I didnt have that problem. ln -s /home/downloads is my friend. ;-) May 09 01:36:25 (that was for flex) May 09 01:36:37 bah May 09 01:36:46 that's cheating May 09 01:36:46 :-) May 09 01:37:02 I wanted to test what a new builder would experience May 09 01:37:45 hehe May 09 01:37:56 hm, radiusd needs /usr/lib/libltdl.so.3 May 09 01:38:01 I'm entitiltled to cheat this time. :-) May 09 01:38:58 ka6sox still around? May 09 01:39:20 he might have fallen asleep... May 09 01:39:22 Maybe the answer is to run a bb fetch then to explicitly checksum the dl directory. I did the 'bb fetch first' thing first time. May 09 01:39:44 libtool... May 09 01:39:54 jbowler, hmmm you mean in the new Makefile? not a bad thought at all May 09 01:40:00 Bummers. May 09 01:40:27 Anyone other than ka6sox have access to a machine shop capable of working 8mm sheet aluminium? May 09 01:41:21 8mm? dang May 09 01:41:30 8mm? Not .8mm? May 09 01:41:36 hehe May 09 01:41:38 that would stop a bullet May 09 01:41:40 8mm indeed. May 09 01:41:57 I'm making a custom bashguard for my chainring. May 09 01:42:08 heh May 09 01:42:12 the last one I made, I used a jigsaw and a drill. May 09 01:42:18 i'm NOT doing that again. May 09 01:42:55 Connecting to ftp.uk.linux.org[195.92.249.252]:21... May 09 01:42:57 ... May 09 01:42:59 ... May 09 01:43:03 looks like it's down :-\ May 09 01:43:15 trying to get mdt May 09 01:43:16 mtd May 09 01:43:18 is that theh story of the day? busted source download sites? May 09 01:43:32 apparently May 09 01:46:14 so I hit Ctrl-C and it got mtd from opesources May 09 01:46:22 but a newbie wouldn't know that May 09 01:52:33 hm, thought the slug had 16MB flash, but it's just 6? May 09 01:53:39 the part is a 8MB part. May 09 01:54:17 3 blocks are used for bootloader and config. May 09 01:57:43 right May 09 02:08:05 heh, I hadn't noticed that the new build system is making 'openslug-packages' May 09 02:08:11 no wonder it's taking so long May 09 02:10:57 The whole system?from guest check-in to the casino and room system?runs on Microsoft Windows XP. And because it is all interconnected, when one goes down, it can all go down. And that happened?disastrously?during the first few days. According to Wynn, the slot-machine system crashed, which also brought down the high-tech room locks. For ten long minutes, until the system came back on-line, guests were both locked out of their rooms and May 09 02:10:57 unable to play the slots. May 09 02:11:09 fscking MS POS crap May 09 02:11:38 there aughta be a law May 09 02:13:00 openslug-packages takes 3 hours on my system (that's parsing all the .bb files - 2293 of them). May 09 02:13:23 jbowler, what's your hardware spec ? May 09 02:15:15 That's without any downloads ('cause I never delete my download directory). The system is a 3GHz Intel Prescott, 1GByte SDRAM (relatively fast), gentoo/reiser4fs (2.6.11-mm3), two 160GByte IDE drives with the swap spread between them and running Raid (1?) May 09 02:16:04 Oh, everythings' compiled -Os -march=prescott May 09 02:17:18 ah, my main build box is an Athlon XP 3200+ (well overclocked 2500+) with 1GB low latency RAM, weak spot is the HD - I need to put one of my 7200rpm ones in there, but originally this was a test install and it's on a 5400rpm drive - noticeably slow May 09 02:17:58 I susually don't delete downloads/sources dir, but in this case I am dling from scratch to see what a newbie would experience May 09 02:18:24 it's building mysql-native now May 09 02:19:47 That's line 3324 of the 6420 line log file. May 09 02:21:15 my bed time I think... May 09 02:27:58 `night jbowler-zzz May 09 02:32:58 so should we set CVS_TARBALL_STASH to oesources in local.conf.template? May 09 02:33:33 what effect would that have ? May 09 02:33:50 seems like oe already goes to oesources as a final resort May 09 02:33:54 dunno - what's the env var to replace the reference to treke? May 09 02:34:29 # Specifies a location to search for pre-generated tarballs when fetching May 09 02:34:29 # a cvs:// URI. Uncomment this, if you not want to pull directly from CVS. May 09 02:34:42 we currently have it pointing at treke May 09 02:35:10 ah, so far I haven't had any problems with CVS dowloads (which is all that would affect I think) May 09 02:36:01 I wonder if they have forwarded whatever.treke.net to oesources ? May 09 02:49:04 03rwhitby * r21 10/conf/local.conf.template: Updated the CVS_TARBALL_STASH URL. May 09 02:49:09 The treke URL is broken. May 09 02:49:18 So I changed it anyway. May 09 02:56:38 cool May 09 02:58:52 we haven't made any Unslung changes since BK went down, right? May 09 03:28:34 I don't think anyone's been able to May 09 03:29:18 wow, openslug-packages takes quite a bit of space May 09 03:40:53 ah, openslug-packages is done :-) May 09 03:44:37 starting another build to get build time without download time May 09 03:46:50 Bah. Package Miau failed on this openslug build May 09 03:47:26 failed how? May 09 03:47:38 The same way your flex failed. May 09 03:47:47 cant d/l it. May 09 03:48:05 ah May 09 03:48:11 for some reason it dl'd for me May 09 03:52:59 oh well, thats where I give up. May 09 03:53:09 I didnt notice it suddenly became 1am. May 09 03:53:24 dyoung: did you get an episode or two? May 09 03:53:38 the first one is gonna finish in "around" 2hrs. May 09 03:53:51 heh May 09 03:53:53 75 min is the current estimate actually. May 09 03:54:49 what rate are you getting? May 09 03:55:01 dyoung you can always download source manually and put it in downloads dir May 09 03:58:13 rwhitby-away, 15KB/s is the current rate. May 09 03:59:06 jacques, the irritating thing is that file is already IN my downloads directory. May 09 03:59:23 hmmmmmm, so what was the error ? May 09 03:59:25 I manually added the md5sum, hopefully that makes it go May 09 03:59:37 you shouldn't have to touch the md5 file May 09 03:59:39 I never do May 09 03:59:43 I never had one. May 09 04:00:00 which indicates I never compiled it for openslug. May 09 04:00:09 it was probably there from unslung package bu ilds May 09 04:00:13 yeah, but it hasn't mattered for me either way - if I don't have one I still don't touch it and it still works May 09 04:00:42 hmm, ig still wants to fetch it May 09 04:00:50 I get a 416. May 09 04:01:01 is it same version ? May 09 04:01:03 "Requested Range not satisfiable" May 09 04:01:07 Yes. May 09 04:01:18 "continued download on this file failed" May 09 04:01:26 same compression scheme (.bz2 vs .gz) ? May 09 04:02:02 .gz all the way around. May 09 04:02:07 well I'll give it once last go. May 09 04:02:22 I blew away the one in downloads May 09 04:03:06 if all else fails, blow it away. :-) May 09 04:03:10 that seemed to go May 09 04:03:31 seems like the one in downloads was not complete and it was trying to continue a download May 09 04:04:51 probably. May 09 04:05:07 it did say something about incompatibility with the -c option May 09 04:05:28 yeah that's what that does May 09 04:07:21 well, the upshot is if this build ever finishes, I can come up with a tick number for jbowler's databse of ticks. I would put it on a pretty early production slug. May 09 04:07:55 ah I was meaning to do that too May 09 04:08:09 if you see me in the morning, remind me to mail lennerts USB/Serial cable. May 09 04:08:14 but hadn't gotten around to figuring out which package I needed and building it May 09 04:08:21 will do May 09 04:08:22 "morning" is around +8 hours. May 09 04:08:25 hh May 09 04:08:26 heh May 09 04:08:39 Thanks, May 09 04:08:45 np :-) May 09 04:50:54 ~seen jp30 May 09 04:51:27 jp30 <~josh@jp30.nslu2-linux> was last seen on IRC in channel #nslu2-linux, 1d 12h 55m 38s ago, saying: 'google are just transitioning to their rightful place as a top level domain; .com's too small for them'. May 09 04:52:15 heh May 09 04:53:37 copperbeech, right before that he said he was going on holiday. May 09 04:54:14 ha ha! jbot didn't catch that one! A lesson to us all about the last thing we say on IRC ! :-) May 09 05:11:22 Which basically means: Don't say *anything* that wouldn't look good as a "famous last words" ;) May 09 05:12:16 :-D May 09 05:12:30 famous last jbot May 09 05:14:38 Anyone got any experience with usb-key wlan? They probably suck signal-strength-wise, but does anyone know how much? May 09 05:49:00 having some strange problems with thttpd... When I try to run it using /opt/etc/inid.d/*thttpd it doesn't run (or stay running) at all, when started manually (without .conf file), it's okay. Is there something wrong with default config file? May 09 05:51:15 which platform? May 09 05:52:30 rwhitby-away: Is there a binutils ipkg for glibc openslug? May 09 05:52:49 * NAiL is trying to set up a native toolchain, but can't find the tools :P May 09 05:54:19 rwithby-away: "which platform?" Was that a question for me? If yes, it's wlhdd (asus). May 09 05:59:22 hey, welcome back stranger! May 09 05:59:59 bob_tm-away: pong May 09 06:00:12 Hey May 09 06:00:21 I will explain what I mean. If you unsling 4.19 to disk1, /.sda1root will be created on the flash and the flash (including the .sda1root) copied to the drive. May 09 06:00:35 But each rc-script seem to disregard diversion if /.sda1root exist May 09 06:01:27 Eg. in /etc/rc.d/rc.1 - first diversion test: if ( [ ! -f /.sda1root ] && [ -f /mnt/sda1/unslung/rc.1 ] && . /mnt/sda1/unslung/rc.1 ) ; then return 0 ; fi May 09 06:01:37 disregard diversion explicitly from disk, cause the disk is already mounted as /, so the scripts in /unslung get run that way May 09 06:02:00 Ah, so I just need to move them from the old conf partition? May 09 06:02:21 did you run the "slingover" script? May 09 06:02:30 I thought we still supported diversion from the conf partition May 09 06:02:37 Nope, just unsling May 09 06:03:01 slingover would have copied that stuff from your old conf partition to the new rootfs May 09 06:03:01 Can you unsling to the conf partition or just the data partition? May 09 06:03:12 default is now data May 09 06:03:24 conf is undocumented for experts with no official support May 09 06:03:33 (should still work) May 09 06:03:54 Why do we still need diversion tests from /mnt/sdXY now? May 09 06:03:57 we've decided to leave conf to linksys and no longer touch it. May 09 06:04:03 Good call May 09 06:04:35 aargh May 09 06:04:41 * NAiL forgot to kill off ntpd May 09 06:04:47 the diversion tests from /mnt/sdXY is so those who use the undocumented unsupported options still have a working system :-) May 09 06:05:16 But then maybe we should mount the disks? May 09 06:05:37 they do get mounted in /etc/rc May 09 06:05:49 Nope, only if you have .sdXroot May 09 06:06:03 those options are for those who keep rootfs on jffs2, but put /opt on the disk. May 09 06:06:19 Hmm - right. May 09 06:06:24 Yes, that's what I did, but then you don't have /.sdXroot and the diversion will not happen May 09 06:06:45 true May 09 06:06:56 ok, not sure how to go forward on that one ... May 09 06:07:12 Can't you mount the disks regardless of /.sdXroot? May 09 06:07:31 back later May 09 06:07:40 bibi May 09 06:09:53 false alarm May 09 06:10:09 ok, mounting disks anyway ... let me think about that. May 09 06:11:29 ok, the trouble is that in Unslung 3.x we were using those flag files to indicate which disks were present. May 09 06:11:45 cause someone could have a vfat disk on sdb, and that would fail the mount May 09 06:12:13 maybe we just take them out altogether, and people have to edit the jffs2 if they want something done. May 09 06:13:17 go back to just /unslung (which could be either jffs2 or external rootfs, but not both) May 09 06:14:19 bob_tm-away: pong May 09 06:14:25 pong.... thinking May 09 06:14:42 Ok - gotta run now, but will think about it. May 09 06:14:46 cool - wans't sure whether you'd gone when I went -away temporarily May 09 06:15:06 I am about to be away. Back in an hour++ May 09 06:15:09 bibi May 09 06:15:25 I won't be here - too late for me tonight. May 09 06:21:27 night all May 09 06:21:52 * rwhitby-asleep wonders if bitkeeper will be working when he wakes .... May 09 06:23:54 real 153m39.620s openslug-packages May 09 06:31:22 ditto. May 09 06:31:25 mine just finished. May 09 06:32:12 cool May 09 07:42:57 My slug is now unslung, thank you all for the work you put into it. Its playtime for me now =) May 09 07:47:39 I get a transferspeed of about 2,5 MB/s with NFS is that normal for a NSLU2 or slow? May 09 07:48:27 D|dge, maybe a little slow..but not unusual May 09 07:54:09 any ideas what the bottle neck can be? May 09 07:57:16 look in the wiki..it tells all May 09 07:57:21 ~nslu2-wiki May 09 07:57:23 methinks nslu2-wiki is the short-hand way of saying "Have a look at the NSLU2-Linux Wiki at http://www.nslu2-linux.org/". May 09 07:58:31 D|dge, there are transfer rates other people have gotten with various methods somewhere on the wiki - I don't recall where May 09 08:05:24 OK, thanx May 09 08:15:11 D|dge: The page is Info.performance - http://www.nslu2-linux.org/wiki/Info/Performance May 09 08:22:30 so mine are the expected values, for writing May 09 08:25:31 I'll try nfs3 later, perhaps it performs better May 09 08:25:35 Thanx and CU May 09 08:25:42 np, later May 09 08:28:00 03bzhou * 10unslung/make/mailman.mk: stripped mailman May 09 08:31:26 any1 good at mail? I built mailman, but need some help in testing it. It would be way cool if nslu2 mailing list is hosted on a slug May 09 08:32:22 at least be able to self host May 09 08:32:24 given the number of people on the list (over 4500) it might be a DEAD slug. May 09 08:34:03 seriously, is the mail traffic that heavy? May 09 08:34:12 it *can* be May 09 08:34:35 if there is a hot and heavy discussion about an issue. May 09 08:35:12 it wouldn't *literally* die...but might get backed up a bit. May 09 08:37:01 maybe i'll just send an email to yahoo group asking for testers. I don't even have postfix setup, and ISP blocks SMTP port, so it's probably a bit difficult for me to test it May 09 08:37:43 sounds like it. May 09 10:52:30 hi May 09 10:52:48 hi there. May 09 10:52:57 any wl-hdd users here? May 09 10:53:47 bleh... irq26: nobody cared May 09 10:54:03 need help configuring it to use the wireless link instead of cable May 09 10:54:20 perlguru, yes I have one and am playing with it. May 09 10:54:42 Larry, when did you get a Wiley? May 09 10:55:11 had one for weeks..using it on a 6km wireless shot outta town. May 09 10:55:27 perlguru, what do you need help on? May 09 10:55:58 need help configuring it to use the wireless link instead of cable May 09 10:56:22 what firmware do you have in it? May 09 10:56:22 communicating router <-> wlhdd over wireless May 09 10:56:30 stock firmware May 09 10:57:11 (no hdd yet) May 09 10:58:15 k May 09 10:58:40 let me bring up a stock one so that I can help you. (I have Oleg's firmware in the other one) May 09 10:59:04 it is possible, right? May 09 10:59:40 yes May 09 10:59:43 (manual is very very not helping) May 09 11:00:41 okay so what you need to do is connect to the default SSID of "WL-HDD and you need to manually setup your IP to 192.168.1.X with .1 as your GW May 09 11:00:59 the default address of the WL-HDD is 192.168.1.1 May 09 11:01:38 from there you can configure the wireless and other stuffs. May 09 11:01:39 emm, 1.220? May 09 11:01:52 192.168.1.220 is my wiley May 09 11:01:59 k May 09 11:02:04 .1.1 is my router May 09 11:02:19 you *can* do this via the wired connection. May 09 11:02:25 (since it will be easier May 09 11:02:43 192.168.1.220 May 09 11:03:18 which page? IP Config ... May 09 11:03:33 pg 7 of the quickstart May 09 11:04:02 i am already on the wiley's webpage May 09 11:04:16 (via wired connection) May 09 12:59:52 ola ! May 09 13:00:26 dyoung: slug arrived this morning :) May 09 13:01:01 jeanfabrice, got your message *just now* May 09 13:01:10 so I'll workon it after lunch May 09 13:01:29 ka6sox: np. I got some internet access troule all the day long. 1.9.2.7-5 is not stable on my wl500g May 09 13:01:39 back to 1.9.2.7-4 May 09 13:02:32 re all May 09 13:05:44 By the way, postfix will not compile on wiley. It depends on a full glibc libresolv and uclibc lacks a lot of function in libresolv. I wonder how it is hard to use a recent uclibc to compile wileyware May 09 13:07:05 uclibc 0.9.19 is so old May 09 13:07:29 I think a lot of apps will compile with newer versions that won't compile with 0.9.19 May 09 13:09:58 jacques: yes. is it really bigger than 0.9.19 so Asus did not use it ? May 09 13:11:33 i think they probably started development in 2003 and never upgraded - it happens a lot May 09 13:11:59 as rwhitby-asleep pointed out, unslung is on glibc 2.2.5 which is equally old May 09 13:12:13 but uclibc has had a lot more advancement in that time May 09 13:12:25 as far as apps compiling against it May 09 13:13:10 hi kitno455 May 09 13:15:48 jeanfabrice, do you want me to hold off then? May 09 13:20:57 0.9.19 3 March 2003 May 09 13:21:18 http://www.uclibc.org/downloads/Changelog May 09 13:23:31 heck, I'd be very surprised if perl builds and tests out with uclibc 0.9.19 May 09 13:23:43 never mind soft-float mode May 09 13:24:29 pcre does not compile so far May 09 13:26:47 I assume wiley firmware has same limitation as unslung? - there are some binary apps which we don't have source for? May 09 13:27:16 0.9.21 is not binary compatible with 0.9.19 for example May 09 13:27:16 wow, ive been gone a couple months, what is wiley (coyote?) May 09 13:28:06 Asus WL-XXX series of wireless routers/HD's May 09 13:28:31 what cpu in wiley? May 09 13:28:42 jacques: akaik, but I'm not totally sure, we don't have binary for wlan if May 09 13:29:10 kitno455, you wouln't believe the names: optware/wiley/Veggie(turnip) May 09 13:29:26 have you ever eaten a turnip? May 09 13:29:29 nasty May 09 13:30:04 mipsel May 09 13:30:32 hmm, i did mips assembly in college, years ago. May 09 13:30:50 BIAB May 09 13:31:21 [g2] was making a turnup page for OpenSlug and dyoung said Turnip? May 09 13:31:36 what is the current status of all these *slug projects? May 09 13:31:38 so now we just refer to it as Veggie. May 09 13:31:45 what does it do? May 09 13:31:53 (veggie i mean) May 09 13:31:59 instructions. May 09 13:33:07 dyoung is the official naming convention maker May 09 14:10:21 I saw some BK changeset going through, is bk fixed now? May 09 14:10:47 are you guys really going to go back before june if it is? May 09 14:10:53 bummer...I was just getting used to Subversion. May 09 14:11:23 SVN, SVN! May 09 14:13:14 I dunno actually. May 09 14:13:38 What I do know is at this rate the repo will rapidly get out of sync. :-) May 09 14:14:37 the SVN one? May 09 14:14:46 both May 09 14:14:58 git May 09 14:15:00 sync requires two, right broadcast enginere May 09 14:15:03 ? May 09 14:37:00 does it make sense to have adduser package (which provide adduser, deluser, addgroup, delgroup) also poviding the 'su' applet (as you can use 'su' to have rc script launch a daemon under an unprivileged account) ? May 09 14:40:42 jeanfabrice, that would be a nice feature. May 09 14:44:12 dyoung: pingu complete? May 09 14:45:58 jeanfabrice, so what is the verdict on -5 builds? May 09 14:48:53 03rwhitby 07 * r1.3246 10openembedded/packages/openslug-init/ (openslug-init_0.10.bb openslug-init-0.10/update-kernel): Added the update-kernel script to openslug-init May 09 14:58:37 was the bitbaking on Athlon 64 systems worked out? May 09 14:58:55 err bitbaking +problem May 09 15:10:45 'nite May 09 15:27:33 rwhitby-away, Yeah, pingu is done; took approx 4.5hrs each. May 09 15:53:57 I notice there are a number of V2.3R25-uNSLUng-able-4.2-alpha entries in the 4.x alpha testers database (including the very latest entry which was added in the last 24 hours). May 09 15:54:16 Is there a really old binary that someone is distributing somewhere? (possibly in Germany) May 09 16:07:25 BK is now fixed, and OpenSlug and Unslung development has moved back to the nslu2-linux.bkbits.net repository. May 09 16:08:04 Note that we will retain the SVN repo, and use it for non-developer source release. It will be updated from the BK repo (and tagged) on each beta binary release. May 09 16:22:43 03bzhou * 10unslung/make/nget.mk: packaged nget May 09 17:21:30 03bzhou * 10unslung/ (make/cogito.mk sources/cogito/Makefile.patch): upstream upgrade from 0.8 to 0.10 May 09 17:38:40 If anyone has any last-minute changes for OpenSlug or Unslung, please speak now (in the next 8 hours). May 09 17:39:54 We are now in feature freeze for both OpenSlug and Unslung, and bug-fixes will only be accepted in the next 8 hours (unless they are critical). We plan to go into full code freeze in 8 hours (when the versions will be bumped, and release processing will begin). May 09 17:40:18 This is also the time to make any required doco updates in the image README and NOTES files. May 09 17:40:26 ack! May 09 17:41:09 ByronT: I updated them last week - as far as I am concerned they just need editorial review. May 09 17:41:10 Hello all May 09 17:41:47 I was hoping someone could help me, I have a very unhealthy slug... May 09 17:41:59 whew... I'm feeling better already, rwhitby May 09 17:42:52 sick slug...oh Slug Dr? May 09 17:42:53 billiamb1: describe the state your slug is in, and how it got there, and we'll see if someone here can help. May 09 17:43:39 just realized this is more the developer room, sorry bout that. May 09 17:44:20 basically, I thought I unslung the NSLU2, and, when I tried to do an ipkg update, I ended up filling the small amount of space on the slug, instead of putting to the HD attached. May 09 17:44:38 jbowler: ping? May 09 17:44:49 now, I can't get into the slug to reflash with Linksys FW,.. when I boot up, the status light blinks and there is no response May 09 17:45:14 billiamb1, do you have access to redboot via serial or telnet? May 09 17:45:17 I can telnet into redboot @ 192.168.0.1 during bootup, but am not clear on how to clear the space necessary to boot up May 09 17:45:22 yes May 09 17:45:41 just unsure of how to make things bette,r and not worse May 09 17:45:41 just reflash using the redboot reflash technique. May 09 17:45:55 that should be on the wiki May 09 17:46:03 ~nslu2-linux May 09 17:46:05 hmm... nslu2-linux is at http://www.nslu2-linux.org May 09 17:46:07 great, I'll do a search now May 09 17:47:56 NAiL: pong May 09 17:47:59 ah May 09 17:48:42 jbowler: I had my slug die on me again, *after* I killed ntpd. I booted, waited a while, killed ntpd, and a few hours later, it died. When booting without ntpd, it stays up... May 09 17:48:43 ok, I found the page I was looking at last night, 'RecoverFromABadFlash', and ... should I be using the SerComm repair option? May 09 17:50:00 I'm not at home at the moment, so I can't start doing a build with netconsole. Can do that tomorrow. May 09 17:50:36 You got any idea what makes the slug panic when ntpd is no longer running? May 09 17:51:37 NAiL: at least it is reproducible. It looked to me as though once the xl205 had been touched as a result of ntpd it kept doing something which eventually resulted in a crash. May 09 17:52:19 Yeah, makes sense. May 09 17:52:59 who's maintaining the x1205 driver, anyways? ;) May 09 17:54:40 billiamb1: Are you in linux? May 09 17:55:09 I belive it's dyoung - the bug is currently assigned to him May 09 17:55:51 ah, ok May 09 17:56:26 I assigned it to him...I thought I saw him say he would look into it sometime this week. May 09 17:56:44 hmm... yes.. he did May 09 17:56:52 on saturday or something May 09 17:56:59 now that you mention it :P May 09 17:57:21 * NAiL has been too busy studying to actually remember any conversations May 09 18:03:14 NAiL, sorry, I have a Linux (suse) laptop set up, which I am about to use to connect to RedBoot, just need to get on and grab sercomm first May 09 18:09:45 billiamb1: Easiest way (in my experience) is to get the upslug util May 09 18:10:32 billiamb1: then 'arping -f 192.168.0.1 && telnet 192.168.0.1 9000', switch on slug, wait, hit ctrl-c once redboot comes up, exit telnet, upslug -u firmware-image May 09 18:11:02 billiamb1: you realise that the SerComm utility is Windows-based, right? May 09 18:11:09 urp... of course forgot... 'enter "upgrade" on the redboot prompt' May 09 18:11:13 upslug works on linux,correct? May 09 18:11:17 yes May 09 18:11:19 billiamb1: correct May 09 18:12:10 rwhitby, ya, just now realizing that, going to try hitting it from both boxes, :), linux first for the redboot, then the Sercomm from the xp-pc... May 09 18:14:12 gotcha May 09 18:14:31 I've been hassling ByronT for a few nights now, he May 09 18:14:39 s been extremely helpful May 09 18:15:20 erasing EEPROM now, looking good May 09 18:25:12 * NAiL assumes all is well and goes to bed May 09 18:25:26 I heard I got assigned x1205. May 09 18:26:31 well...you said you would "look at it sometime next week" May 09 18:26:53 :) May 09 18:27:35 translation, "I'll look at it when I get around to it". :-) May 09 18:27:48 there is *no* timeline. May 09 18:33:50 Linksys firmware is back =) May 09 18:34:03 billiamb1, whew. May 09 19:12:45 03bzhou * 10unslung/ (make/cogito.mk sources/cogito/Makefile.patch): fixed a lib path problem, bumped IPK_VERSION May 09 19:16:37 can anyone verify whether /usr/bin/env is symlink to /opt/bin/env on unslung? May 09 19:20:22 eno: I can verify in about 45 minutes May 09 19:28:01 ka6sox-office, thx May 09 19:40:48 <[g2]> any debian guys know how to apt-get install psyco ? May 09 19:52:32 can anyone else verify that svn.berlios.de is down May 09 19:52:33 ?? May 09 19:53:34 i am currently unable to get bitbake May 09 19:54:07 .... May 09 19:55:24 hello May 09 19:55:29 can anyone help me May 09 19:55:32 ?? May 09 19:56:24 * james slaps Christopher-bbl around with a small 50lb Unix Manual May 09 19:56:52 the BK repo is back up May 09 19:57:19 friend of yours? May 09 19:57:20 i still can't connect to svn May 09 19:57:29 is it just me? May 09 19:57:35 so you don't need the svn repo May 09 19:57:44 he needs the svn repo to get bitbake May 09 19:57:50 ah May 09 19:58:05 yes! i do, i do May 09 19:58:13 Well, I did a fresh pull of the bitbake files around 18 hours ago, it worked then. May 09 19:58:15 can you guys help? May 09 19:58:28 <[g2]> I pulled earlier today May 09 19:58:29 i just keep getting connectioin refused May 09 19:58:34 I can browse the svn repo May 09 19:58:38 what specific command are you using? May 09 19:59:01 just the command listed on the tutorial "svn co svn://svn.berlios.de/bitbake/trunk/bitbake" May 09 19:59:45 "[root@PC-1 slug]# svn co svn://svn.berlios.de/bitbake/trunk/bitbake May 09 19:59:46 svn: Can't connect to host 'svn.berlios.de': Connection refused May 09 19:59:50 I get a 'connection refused' too May 09 20:00:09 ahhh, so it is just a waiting game.. May 09 20:01:26 <[g2]> james, I've got a tarball from a pull that's bzip2'd May 09 20:01:44 <[g2]> it's only 99K May 09 20:01:53 mmmm, i am a bit of a novice.. May 09 20:02:22 Also an 'svn up' on my existing bitbake svn tree (well, a copy of it) fails with connection refused too. May 09 20:02:29 <[g2]> personally, I'd check the doco out and wait for svn to come back up May 09 20:02:35 i am stumbling my way around just using the tutorial May 09 20:02:49 okay then, adavice taken May 09 20:03:07 have a good day, no doubt i will be back! May 09 20:03:09 seeya May 09 20:03:12 thanks again May 09 20:03:16 <[g2]> np May 09 20:05:11 [g2]: apt-get install python-psyco ? May 09 20:07:42 <[g2]> eno, We have a winner ! May 09 20:07:49 * [g2] hugs eno May 09 20:08:55 * [g2] is actually building on the laptop running Knoppix May 09 20:09:23 i c. When in doubt, you can search by 'apt-cache search psyco' May 09 20:14:06 <[g2]> eno, THX... just like emerge -s May 09 20:19:50 ~seen NAiL May 09 20:19:51 nail is currently on #nslu2-linux. Has said a total of 23 messages. Is idling for 2h 8m 34s May 09 20:21:35 NAiL: I'm going to run a slug without added NTP for a while. Worth double checking that it really is ntpd which sparks the problem. May 09 20:24:53 can someone tell me how to open nano on the slug? I did an 'ipkg install nano', but can't seem to run it, to make some openssh changes May 09 20:25:29 03bzhou * 10unslung/ (Makefile make/nget.mk sources/nget/example.ngetrc): nget ready for testing (cross) May 09 20:26:22 billiamb1, what's the output if you type 'type nano' May 09 20:26:40 nano: No such file or directory May 09 20:27:22 i got: May 09 20:27:23 $ type nano May 09 20:27:23 nano is /usr/bin/nano May 09 20:27:41 ls -l /usr/bin/nano May 09 20:27:43 lrwxr-xr-x 1 root root 9 2005-04-10 10:55 /usr/bin/nano -> /bin/nano May 09 20:29:01 sorry, I take it back. Typing in the wrong window, and wrong machine May 09 20:29:02 are you guys on a nslu2? May 09 20:29:03 are you guys on a nslu2 or WL5xx? May 09 20:29:42 I'm trying to follow the 'UseOpenSSHForRemoteAccess', and it states to make some changes, and recommends nano, so I did an 'ipkg install nano', I just did it again, and was told "Package nano (1.2.4-2) installed in root is up to date. Nothing to be done. May 09 20:29:50 on an NSLU2 May 09 20:30:44 i got: May 09 20:30:46 type nano May 09 20:30:46 nano is /opt/bin/nano May 09 20:31:10 and just installed nano May 09 20:31:39 maybe try 'ipkg remove nano; ipkg install nano' May 09 20:32:36 hmm, if I type 'type nano', I get the above error, if I "cd /opt/bin", that's fine, if I then "cd /nano", I get "can't cd to /nano/" May 09 20:33:26 now, after the reinstall, nano is listed on the 'ls -l' for /opt/bin, but still states 'can't cd to nano' when I 'cd /nano' May 09 20:33:34 uh..... May 09 20:33:43 of course you cant. May 09 20:33:44 could be a PATH problem, what if you use the full path to start nano: /opt/bin/nano May 09 20:33:59 that works! May 09 20:34:00 what exactly are you trying to do May 09 20:34:36 I'm trying to follow the 'UseOpenSSHForRemoteAccess'.. May 09 20:35:55 and, it's stumping me May 09 20:36:14 what step are you on? As far as I can see it never tells you to "cd /nano" May 09 20:36:22 because that wont do anything and wont work. May 09 20:37:00 ok, I guess that's my issue, it states that I need to add an option into the OpenSSH configuration script, and the write recommended nano.. May 09 20:37:24 I'm just a bit novice, is my issue May 09 20:37:51 so, if i misread anything here, please don't hurt me :) May 09 20:38:31 billiamb1, you can use /opt/bin/nano for now, but you need to fix your PATH. If you use /opt/bin/bash as your shell, /opt/etc/profile sets the default PATH so that you can find nano without the full path May 09 20:38:39 I'm assuming youre on step 9 May 09 20:38:56 dyoung: yes, I am May 09 20:39:15 although May 09 20:39:46 my public key, as created with the PuTTY Key Gen, does not end with a someone@hostname, .. it ends with rsa-key-20050509 May 09 20:40:01 so, I may have messed up on step 8 to be honest May 09 20:40:16 eno, that makes sense, that's my next mission I guess May 09 20:46:29 so, on step 9, it states to change "PasswordAuthentication" to 'no' in the "sshd_config", I don't have that currently listed...I can add it in, but was concerned about it's not being there... is that an issue? May 09 20:53:46 billiamb1, my /opt/etc/openssh/sshd_config has that line commented out, so I guess it defaults to 'no'. And this is a developer discussion channel, so don't expect always get answer here May 09 20:54:32 understood, no worries.. it's getting late, I'll jump on the general channel tomorrow, see if I can wrassle this all out. May 09 20:54:42 thanks a lot for the help all! May 09 20:55:01 just glad I was able to get it back to a slug May 09 22:32:58 03bzhou * 10unslung/sources/vim/configure.in.patch: patch for src/configure.in to make vim cross build May 09 22:36:07 jbowler-away: I'm running without ntp now. I'll check status when I get back home. If the slug is still running, I'm 95% certain it's ntpd causing the crashes by basically being started and running a while before being killed. May 09 22:45:06 NAiL: yes, that what I think too. Mine is still going. Probably easier to just go in there and debug it ;-) May 09 23:12:48 03cdoban * 10unslung/make/asterisk.mk: builds from CVS May 09 23:13:19 Thanks cdoban! May 09 23:13:52 I can see it now...the Slug becomes my PBX...cool. May 09 23:14:49 03cdoban * 10unslung/sources/asterisk/ (Makefile.patch codecs.gsm.Makefile.patch): builds from CVS May 09 23:18:01 hi. I'm hoping to get some quick advice. I have a minimal nslu2 I'm hacking on that can't have an external hard drive (using the USB's for something else). May 09 23:18:32 I need to mangle text like I normally would with perl. what kind of minimal scripting language can/should I use? .. I'm on the verge of just writing my own in C. May 09 23:18:50 03cdoban * 10unslung/sources/asterisk/ (4 files): remove old patches May 09 23:20:51 happ what kind of devices are you using? May 09 23:21:46 I'll be using the NSLU2 as a ethernet/usb-serial gateway as the backchannel for a wireless sensor network testbed May 09 23:22:25 Aha, so you can use USB1.1. You should be able to slap a USB1.1; and it should work for 2 devices. May 09 23:22:41 freeing one of the onboard usb ports for a flashstick. May 09 23:22:54 yes. the ehci driver was a serious pain in the ass. everything is better now that I'm just loading up the ohci driver May 09 23:23:00 YMMV though. There have been mixed results. May 09 23:23:32 well, we'll be selling these in bulk, so requiring a flash stick per NSLU2 isn't an option just to include, say, perl May 09 23:23:47 For myself I've emperically determined that 2 PL2303's work behind a hub with a flash stick in the other port. May 09 23:23:59 Hmm, ok. May 09 23:24:23 I think youre not gonnabe able to fit perl even if you fill the flash. May 09 23:25:09 happ, have you condered other embedded devices for your application? May 09 23:25:27 one immediately comes to mind. lemme find the link... May 09 23:25:57 we saw the kuro box. but the nslu2 is nice because it's so cheap :). don't convince me on another, we already ordered 300 May 09 23:26:11 (of the nslu2's) May 09 23:26:28 http://www.embeddedarm.com/epc/ts7250-spec-h.html May 09 23:26:33 we have everything working, I just need to mangle some output for cgi... May 09 23:26:40 WOW May 09 23:26:47 okay, you win. May 09 23:27:38 you might be able to squeeze it into unslung if you dont ipkg update, and blow away the samba related files. May 09 23:29:25 hrmph. I was hoping maybe even for lib pcre, but I think I'd need a native compiler for build it, and right now we're working just with cross compilers (since we don't have any hd's hooked up) May 09 23:29:59 have you tried any of this with openslug yet? May 09 23:30:11 2.6 kernel. May 09 23:30:16 well, unslung, not openslug per se May 09 23:30:53 we're using the ftdi usb/serial chip. my experience is that at least for that chipset, usb in 2.6 is more unstable than 2.4 May 09 23:31:18 theres a pcre mk file for unslung. May 09 23:31:44 ah, I thought I looked, I go look again May 09 23:31:45 its in the CROSS section, so it will build cross. May 09 23:32:42 happ, have you considered making youro own custom rootfs specific to your application? May 09 23:33:07 yeah, we'll be getting there eventually.. we're already hacking the rootfs somewhat. May 09 23:33:20 stability before features, though, so that's slated for v1.1 :) May 09 23:33:54 you should be able to hack-n-slash the unslung kernel stuff you glean from the bk repo. May 09 23:38:51 ya, I have to say that unslung is extremely nice May 09 23:41:57 Thanks. May 09 23:42:21 Your application sounds pretty cool. May 09 23:42:41 I've done something similar but with a expensive Digi product. May 09 23:45:34 well heck, if you like hacking on embedded systems, you might think our motes are cool http://moteiv.com/ and the "operating system" that runs on them http://tinyos.net/ May 09 23:46:17 10k ram, 48k code :) May 09 23:47:00 thanks for the pointer on pcre, I have it compiled and ready to go May 09 23:50:02 Cool! May 09 23:50:21 hope pcre works out for you. May 09 23:50:31 ya, it should, thx May 09 23:51:44 you guys hve some pretty neat stuff. May 09 23:51:53 are these boards of your own design? May 09 23:53:47 yeah, not me, our other guy Joe did them.. he's about to finish up his PhD May 09 23:54:34 we all work at Berkeley, too.. and that's about the 4th design, give or take, hence moteIV May 09 23:59:05 thats very cool stuff. :-) **** ENDING LOGGING AT Mon May 09 23:59:56 2005