**** BEGIN LOGGING AT Mon Dec 18 03:00:05 2006 Dec 18 03:35:54 03bzhou * r4744 10optware/trunk/make/py-mercurial.mk: py-mercurial: 0.9.2 -> 0.9.3 Dec 18 03:38:10 03bzhou * r4745 10optware/trunk/make/py-paste.mk: py-paste: 1.0 -> 1.0.1 Dec 18 04:57:38 for APEX there are two boot loaders: nslu2 and openslug - what is the diff? Dec 18 04:59:09 config differences Dec 18 04:59:29 kernel/ramdisk/rootfs, etc. Dec 18 05:00:29 rwhitby thkx - no mention of it on wiki or on APEX page of course :-( Dec 18 05:00:46 well, it is pretty bleeding edge at the moment Dec 18 05:01:07 even more bleeding than fatslug eh Dec 18 05:02:36 rwhitby so is it too risky to experiment with 2nd stage bootloader and then debian LE install? Dec 18 05:03:11 and you just invoked rule #5 Dec 18 05:03:43 no, second stage bootloader is pretty safe, as long as you have serial Dec 18 05:04:58 yeah when/if I get this working, I'll update wiki, provided someone else didn't spam it. It's really amazing how much talent goes into abusing something versus something constructive. Dec 18 05:05:28 yeah I saw that in the wiki... install the APEX into RAM and then run. I got that to work correctly. Dec 18 05:06:49 I just need to figure out the memory addresses and the like so that I don't screw up redboot (!!) Dec 18 05:07:43 I would suggest you use slugimage and upslug2 Dec 18 05:08:32 and start with the apex config in the debian apex-nslu2 package Dec 18 05:09:44 Already got debian LE per http://www.cyrius.com/debian/nslu2/ and prefer that much moreso over unslug, etc. that works quite well. I was hoping that the newer debian install would be a little easier than my previous debian install (lots of problems with kernel modules but thanks to forum - got 'em worked out). Dec 18 05:12:38 so what you mentioned previously - /dev/mtd0 -> /dev/mtd5 - is that contiguous? I.e., "cat mtd0 mtd1 mtd2 mtd3 mtd4 mtd5 >wholeflash.img" Dec 18 05:12:57 yep Dec 18 05:13:15 you can then unpack that with slugimage to confirm Dec 18 05:13:21 I see... thks... any diff b/w mtd0 and mtdblock0 Dec 18 05:15:15 for reads, I don't think so. Dec 18 05:16:13 03bzhou * r4746 10optware/trunk/make/ (py-protocols.mk py-ruledispatch.mk): py-protocols, py-ruledispatch: appeases autobuild Dec 18 05:16:57 gotcha... and presumably mtd0ro is read-only as the name suggests. It just seems weird how a 4MB flash is broken up into 6 parts (0->5) Dec 18 05:17:26 I would think it would be a power of 2 (or at least that would be how I would do it) Dec 18 05:20:30 oops I meant 8MB flash, I think Dec 18 05:21:37 cat /proc/mtd Dec 18 05:22:11 03bzhou * r4747 10optware/trunk/make/py-sqlobject.mk: py-sqlobject: renamed py25 /opt/bin/* Dec 18 05:22:19 ah thanks d'oh I knew that but forgot about it until you mentioned it Dec 18 05:22:39 NSLU2:/dev# cat /proc/mtd Dec 18 05:22:39 dev: size erasesize name Dec 18 05:22:39 mtd0: 00040000 00020000 "RedBoot" Dec 18 05:22:39 mtd1: 00020000 00020000 "SysConf" Dec 18 05:22:39 mtd2: 00100000 00020000 "Kernel" Dec 18 05:22:39 mtd3: 00020000 00020000 "Ramdisk" Dec 18 05:22:41 mtd4: 00660000 00020000 "Flashdisk" Dec 18 05:22:44 mtd5: 00020000 00020000 "FIS directory" Dec 18 05:25:46 what did we do before /proc ... Dec 18 05:29:38 scratch what I said before about power of 2 - didn't quite think about it in regards to partitions of mtd Dec 18 05:44:49 03bzhou * r4748 10optware/trunk/make/py-sqlobject.mk: py-sqlobject: fixed typo Dec 18 05:47:09 03bzhou * r4749 10optware/trunk/make/py-sqlobject.mk: py-sqlobject: fixed typo Dec 18 05:47:11 reading Intel datasheet 252480-006 apparently the max memory supported is indeed 256MB and it even supports bank interleave. Wonder if this is enabled and if anyone has ever gotten this to work? Dec 18 05:48:04 http://osuosl.org/files/rack1-full-11092006.jpg - NSLU2-Linux first in the list :-) Dec 18 05:54:22 03bzhou * r4750 10optware/trunk/make/py-ruledispatch.mk: py-ruledispatch: too many mistakes tonite, I should go to bed Dec 18 07:46:45 make menuconfig Dec 18 07:46:50 oops Dec 18 08:33:32 Your display is too small to run Menuconfig! Dec 18 08:33:32 It must be at least 19 lines by 80 columns. Dec 18 08:33:38 :-D Dec 18 08:37:57 wow Dec 18 08:38:02 never gotten that one before :P Dec 18 12:49:23 Does anyone happen to know what the default dropbear equivalent filename is for the typical openssh ~/.ssh/identity or ~/.ssh/id_dsa keys ? Dec 18 12:50:19 i am trying to setup key-based authentication ... and when using the -i my_identity_file command line switch ... the key-based authentication works ... but if possible i would like to not Have to constantly provide that -i parameter Dec 18 14:39:46 03oleo * r4751 10optware/trunk/ (5 files in 2 dirs): transmission: r1226->r1255 - merge NAT traversal Dec 18 14:41:52 03bzhou * r4752 10optware/trunk/make/nginx.mk: nginx: 0.5.3 -> 0.5.4 Dec 18 14:58:56 03bzhou * r4753 10optware/trunk/make/py-pastedeploy.mk: py-pastedeploy: fixed endless autobuild of this package Dec 18 17:35:19 ~seen NAiL Dec 18 17:35:32 nail is currently on #oe #nslu2-linux #openslug. Has said a total of 13 messages. Is idling for 8h 42m 54s, last said: 'never gotten that one before :P'. Dec 18 17:36:55 NAiL ! I am restructuring buildroot to easily accomodate different uClibc configs. Tou should fix ts101 port accordingly. Dec 18 17:46:36 cool Dec 18 17:46:43 well, kinda cool :P Dec 18 17:49:51 I meant, yous Dec 18 17:49:56 oops. Dec 18 17:50:15 I meant, YOU hould fix ts101 port accordingly. Dec 18 17:50:22 But this will be aseier Dec 18 17:50:53 Whata bogus keyboard on this laptop:) Dec 18 18:05:12 run trackbar Dec 18 18:10:42 morning VoodooZ Dec 18 18:10:57 oleo: yeah, I assumed I had to fix it ;) Dec 18 18:22:56 morning NAiL Dec 18 18:23:29 I haven't been keeping up with the irc lately. What's big and new? Dec 18 18:24:48 NAiL , I wanted to fix it but I doubt that it worked ever. Some hunks are not aligned well. Dec 18 18:33:10 now I see that you did not hav your gcc-uclibc-3.x.mk.patch Dec 18 18:33:25 so mistake is mine. Dec 18 19:52:57 Jim's article is up: http://www.smallnetbuilder.com/nas/nas-how-to/hacking-linksys-nslu2-debian/ Dec 18 20:02:28 03bzhou * r4754 10optware/trunk/make/ (libevent.mk memcached.mk nylon.mk tor.mk vdr-mediamvp.mk): libevent: 1.2 -> 1.2a Dec 18 20:19:33 eno: vdr-mediamvp sounds interesting, do you want to use vdr to stream to the hauppauge mvp? Dec 18 20:24:00 gda_: i'm not working on vdr-mediamvp, just upgraded libevent, and it was using libevent1.2.so Dec 18 20:25:27 eno: sorry misunderstood, do you know who is using it? Dec 18 20:26:49 not sure, the initial check-in was done: dom | 2004-10-14 Dec 18 20:26:59 probably ask on the list Dec 18 21:44:53 hello! Dec 18 21:45:00 I stole your bash script to spindown my external usb drive from http://www.nslu2-linux.org/wiki/FAQ/SpinDownUSBHarddisks - and I made some modifications so that it actually works. anyone can tell me to whom I can write a email and tell them about it? ;) Dec 18 21:58:01 pat: write it in the wiki ;) Dec 18 21:58:29 thats a very good idea. should I hack the password too? ;) Dec 18 21:58:47 d'oh.. i forgot it's password protected :) Dec 18 22:00:06 dwery: see the bottom of that page Dec 18 22:00:27 pat: see the bottom of that page. It was also posted to the mailing list yesterday. Dec 18 22:01:17 ah, there it is, thanks. Dec 18 22:01:42 for some reason, that page regularly gets wiki-spam. we're doing an experiment to see if a password will stop that. Dec 18 22:44:54 Hi, I have added kissdx to the package list at http://www.nslu2-linux.org/wiki/Optware/Packages. Can someone with a password approve the link to http://kissdx.vidartysse.net/ ? Dec 18 22:52:45 vtysse: done Dec 18 22:53:25 Hey thanks, NAiL :-) Dec 18 22:53:39 np Dec 18 23:06:09 03oleo * r4755 10optware/trunk/make/sane-backends.mk: sane-backends: switch to CVS distribution Dec 18 23:25:55 03bzhou * r4756 10optware/trunk/make/ (py-paste.mk py-pastedeploy.mk py-pastescript.mk): py-paste*: 1.0 -> 1.1 Dec 19 01:09:50 03osas * r4757 10optware/trunk/make/asterisk.mk: asterisk: 1.2.13 -> 1.2.14 Dec 19 01:28:51 03osas * r4758 10optware/trunk/make/asterisk.mk: asterisk: reset ASTERISK_IPK_VERSION for asterisk 1.2.14 Dec 19 01:38:10 * rwhitby flashes openwrt built for the nslu2 ... Dec 19 01:46:05 openwrt has support for the npe? Dec 19 01:46:23 it does now Dec 19 01:46:31 (using the new open source driver) Dec 19 01:47:02 I bet you could fit two copies of openwrt in the flash on the NSLU2! Dec 19 01:47:25 and four on the nas100d or dsmg600 Dec 19 01:47:46 well, it boots the kernel, but doesn't get much further: http://pastebin.ca/284454 Dec 19 01:50:02 Wow - not bad! Does it not like the rootfs, then? Dec 19 01:51:03 dunno. the same build works on the nas100d. Dec 19 01:54:51 ok, so the first problem is the image - at 0x1600000 I don't have the SerComm header. I need to update slugimage to always put that in if the kernel is less than 1MB and you don't ask for a ramdisk. Dec 19 01:55:37 but we don't have a ramdisk in the standard firmware currently... Dec 19 01:55:48 (do we?) Dec 19 01:55:58 well, we have an empty erase block Dec 19 01:56:28 (it's required, cause Linksys RedBoot copies according to the length header at that address. Dec 19 01:56:34 Ah - is this perhaps because we don't really have fis partitions on the NSLU2? Dec 19 01:57:00 so if it sees a jffs2 partition at 0x160000, it will interpret the first four bytes as a length to copy, and clobber everything else. Dec 19 01:57:14 (that's why I had to manually load apex in that pastebin) Dec 19 01:57:27 but that shouldn't affect the kernel loading the jffs2 as rootfs. Dec 19 01:57:51 so it doesn't explain the failure of the rootfs, only the failure of apex to boot automatically from flash. **** ENDING LOGGING AT Tue Dec 19 03:00:00 2006