**** BEGIN LOGGING AT Sat Mar 12 23:59:56 2005
Mar 13 01:07:36 <pTweety> hi
Mar 13 01:07:58 <pTweety> can someone help me to increase the samba performance?
Mar 13 01:08:19 <pTweety> currently it seems to be a bit slow ~250KB/s
Mar 13 01:08:39 <pTweety> does samba v3 increase he speed
Mar 13 01:08:41 <pTweety> ?
Mar 13 03:27:08 <rwhitby> hey hanjo
Mar 13 03:27:14 <rwhitby> got an image working yet?
Mar 13 03:27:20 <hanjo> :)
Mar 13 03:27:24 <hanjo> no not yet
Mar 13 03:27:30 <hanjo> i.e
Mar 13 03:27:44 <hanjo> your image looks ok on the outside, but I can not get in the box :)
Mar 13 03:28:01 <hanjo> and my new build is blocked on the CVS
Mar 13 03:28:11 <rwhitby> hmm - that was the latest build on the official build machine.  but no-one has tested that exact image, just images built from the same source.
Mar 13 03:29:26 <hanjo> exact same sources?
Mar 13 03:30:02 <hanjo> I don't know what can be wrong. 
Mar 13 03:30:40 <hanjo> The IP and the GW are properly set. Without changing anything and just reflashing linksys I can telnet inside
Mar 13 03:31:12 <hanjo> but with that image, the slug is not recognized on the network
Mar 13 03:31:25 <hanjo> even when I am directly connected with a cross-over cable
Mar 13 03:33:50 <hanjo> but I am glad that I have my build environment sorted out now
Mar 13 03:34:21 <hanjo> that patch issue was really hard to track
Mar 13 03:38:30 <rwhitby> hanjo: I will send you the latest image I built locally.  I haven't tested that one either, but it should work :-)
Mar 13 03:41:15 <hanjo> rwhitby: I have to be doing something wrong then...
Mar 13 03:41:47 <rwhitby> maybe not - the previous one might have had the problem where modules were in two dirs
Mar 13 03:42:38 <hanjo> the 11.2 issue?
Mar 13 03:42:39 <rwhitby> just checked - it did have that problem
Mar 13 03:42:42 <rwhitby> yes
Mar 13 03:42:53 * hanjo is happy
Mar 13 03:42:58 <rwhitby> so this new one I am uploading for you should be better
Mar 13 03:43:17 <hanjo> btw, any feedback from SF on the CVS mess?
Mar 13 03:43:25 * rwhitby starts a clean build on the official machine ....
Mar 13 03:43:39 <rwhitby> didn't know about it
Mar 13 03:43:40 <hanjo> I should not have cleaned all the sources
Mar 13 03:47:01 <rwhitby> I told you not to :-)
Mar 13 03:47:10 <rwhitby> 25789a6812c83b42b4abd64f1355c8ac  openslug-nslu2-20050312041257.flashdisk.img
Mar 13 03:47:15 <rwhitby> grab it now
Mar 13 03:49:25 <hanjo> done
Mar 13 04:42:12 <rwhitby-away> Unslung/WL500g has been born.
Mar 13 04:43:15 <rwhitby-away> I have convinced Oleg (author of the Asus wl500g wireless router custom firmware) to use the Unslung package repository template.mk format and /opt installation location for the packaging system for that firmware.
Mar 13 04:43:44 <rwhitby-away> 5 lines change to top-level Makefile, and 90 packages build with no changes to package.mk files.
Mar 13 04:44:42 <rwhitby-away> "your buildsystem rocks:
Mar 13 04:44:43 <rwhitby-away> builds/gift_0.11.8.1-1_mipsel.ipk  builds/tcpdump_3.8.3-1_mipsel.ipk
Mar 13 04:44:43 <rwhitby-away> builds/hdparm_5.9-1_mipsel.ipk     builds/zlib_1.2.2-1_mipsel.ipk
Mar 13 04:44:43 <rwhitby-away> builds/strace_4.5.8-1_mipsel.ipk"
Mar 13 05:02:34 <rwhitby-away> night all
Mar 13 06:51:37 <thaylok> anyone awake? :)
Mar 13 06:52:17 <thaylok> my original firware slug have gone mad :(
Mar 13 06:53:05 <[g2]> ~nslu2-rules
Mar 13 06:53:06 <jbot> well, nslu2-rules is 1) Those who ask the questions, update the wiki, 2) Those who complain about the doco, update the wiki, 3) Friends don't let friends flash custom firmware without confirmed RedBoot access, 4) Friends don't let friends flash custom bootloaders without confirmed JTAG access.
Mar 13 06:54:09 <[cc]smart> thaylok: flash it 
Mar 13 06:55:00 <thaylok> hehe.. my first thought too, but can't get http access no mather what I do :(
Mar 13 06:55:26 <[cc]smart> http://www.nslu2-linux.org/wiki/HowTo/TelnetIntoRedBoot
Mar 13 06:56:28 <[cc]smart> http://www.nslu2-linux.org/wiki/Main/UpSlug
Mar 13 07:29:25 <evil-dna> yo
Mar 13 07:29:35 <evil-dna> NSLU2 miught be my next toy!
Mar 13 07:39:12 <evil-dna> how much flash memory does the NSLU2 have?
Mar 13 07:40:27 <[cc]smart> 16M?
Mar 13 07:40:56 <[cc]smart> but well, what u usually do is add a usb stick or HD :)
Mar 13 07:41:59 <evil-dna> hm right.. it says 8MB on the site
Mar 13 07:42:09 <evil-dna> tho there are only 2 USB connectors :)
Mar 13 07:42:35 <[cc]smart> if you choose openslug, chances are that means you can connect two usb hubs
Mar 13 07:42:47 <evil-dna> hm.. right
Mar 13 07:43:01 <evil-dna> openslug would support a loop filesystem?
Mar 13 07:43:09 <[cc]smart> lemme check... one sec
Mar 13 07:43:27 <evil-dna> ie main fs on the USB2 drive = fat32, then an ext2 fs nested in a file with additional tools etc.. for linux?
Mar 13 07:44:08 <[cc]smart> kernel-module-loop_2.6.11.2-r0_nslu2.ipk
Mar 13 07:44:13 <[cc]smart> so i would think so
Mar 13 07:44:15 <evil-dna> hehe nice
Mar 13 07:44:26 <[cc]smart> but i formatted my USB stick ext3
Mar 13 07:44:40 <evil-dna> what id love to do if i get this device is to run a bittorrent clent on it.. think its possible?
Mar 13 07:45:25 <evil-dna> that would truly kick ass
Mar 13 07:46:07 <[cc]smart> i'm not the one who could tell you how to do it, but check the mail archives
Mar 13 07:46:25 <[cc]smart> i'm quite sure there were several who have set it up
Mar 13 07:54:58 <evil-dna> s*$! i burnt myself with hot wateer
Mar 13 07:55:26 <evil-dna> thing splashed on my face as i dropped the glass
Mar 13 07:55:38 <evil-dna> it wasnt boiling thankfully
Mar 13 07:56:14 <evil-dna> no cappuccino for me :(
Mar 13 07:57:06 <[cc]smart> hot water can be very dangerous. worst case if you spill it onto clothes that suck it up and hold it right to your skin.
Mar 13 07:57:17 <[cc]smart> that's fun
Mar 13 07:58:36 <evil-dna> :(
Mar 13 07:58:52 <evil-dna> it happened to my sister once wwhile cooking pasta :(
Mar 13 07:58:57 <evil-dna> took her to a doc
Mar 13 07:59:14 <evil-dna> she had a brown stain on her belly for a while
Mar 13 08:00:43 <evil-dna> ya anyway i started to write a pure C BT client last client but project is stalled.. i might resume it some time later
Mar 13 08:00:57 <evil-dna> last summer*
Mar 13 08:01:22 <[g2]> I think some ppl are running ctorrent
Mar 13 08:07:16 <evil-dna> ya i found libbt
Mar 13 08:15:48 <thaylok> ugly little slug!
Mar 13 08:22:15 <thaylok> anyone ever had a problem with the webinterface on an original linksys firmware?
Mar 13 08:28:02 <caplink811-away> thaylok: See http://www.nslu2-linux.org/wiki/HowTo/ResetSysConf ... after Translation point 1 to 3
Mar 13 08:28:12 <caplink811-away> ~nslu2-general
Mar 13 08:28:14 <jbot> #nslu2-general is the IRC channel specifically set up for end-user questions about the NSLU2 device. Questions that are not related to the nslu2-linux development projects should not be asked in #nslu2-linux, but should be asked in #nslu2-general instead.
Mar 13 08:30:42 <thaylok> caplink811-away: thanks.. :)
Mar 13 08:42:05 <CIA-12> 03ingeba * 10unslung/sources/libxml2/configure.patch: Remove local library search paths
Mar 13 08:43:18 <CIA-12> 03ingeba * 10unslung/sources/libxslt/libtool.patch: Patch configure instead
Mar 13 08:44:17 <CIA-12> 03ingeba * 10unslung/sources/libxslt/configure.patch: Remove local library search paths
Mar 13 08:44:55 <CIA-12> 03ingeba * 10unslung/make/ (libxslt.mk libxml2.mk): Patches to compile cross
Mar 13 09:01:30 <jp30> bob_tm, thanks for the work on libxslt
Mar 13 09:01:30 <thaylok> caplink811-away: that fixed it :) big thanks!
Mar 13 09:01:48 <bob_tm> jp30: No problem 
Mar 13 09:03:08 <CIA-12> 03jp30 * 10unslung/make/freetype.mk: freetype: follow the -config script in STAGING/bin convention
Mar 13 09:09:55 <CIA-12> 03ingeba * 10unslung/make/gift.mk: Remove deletion of staging/opt/share and man
Mar 13 09:28:38 <bob_tm> jp30: Added http://www.nslu2-linux.org/wiki/Unslung/FirmwareWishList and made link from main Unslung page
Mar 13 09:29:44 <jp30> thanks, bob_tm, i'll take a look
Mar 13 09:29:48 <CIA-12> 03jp30 * 10unslung/ (6 files in 3 dirs): deprecate libiconv
Mar 13 10:07:41 <morale> this is getting annoying having to reboot every 2 minutes 
Mar 13 10:10:31 <morale> could it be a problem in the 2.4.29 kernel which is causing this lockup/forced reboot on v2.2 wrt54g's ?
Mar 13 10:25:41 <evil-dna> this is #nslu2-linux :)
Mar 13 10:26:13 <evil-dna> morale: so you managed to install experimental after all? nice :D
Mar 13 10:26:38 <evil-dna> i only have batbox now
Mar 13 10:26:47 <morale> batbox?
Mar 13 10:26:51 <evil-dna> ya
Mar 13 10:27:04 <evil-dna> it's a ram-only install for wrt54g
Mar 13 10:27:09 <evil-dna> very basic tho
Mar 13 10:27:15 <morale> im going to exchange my wrt54g for a wrt54gs 
Mar 13 10:27:22 <morale> the wrt54gs doesn't crash does it?
Mar 13 10:27:37 <evil-dna> you really need a gs?
Mar 13 10:27:46 <evil-dna> speebooster only works with linksys products
Mar 13 10:28:13 <morale> i just need something that doesn't freeze up 
Mar 13 10:30:55 <evil-dna> then just wait until experimental gets included in the official openwrt
Mar 13 10:31:02 <evil-dna> it will be very soon im sure
Mar 13 10:31:28 <evil-dna> or read the forums to see how other people did.. it seems to work well if you do it right
Mar 13 10:31:31 <morale> im going to get a wrt54gs just so i have the extra 16mb of memory too
Mar 13 10:31:46 <evil-dna> im not sure its really required
Mar 13 10:32:04 <morale> hmm..
Mar 13 10:32:08 <evil-dna> i would just keep my wrt54g.. why spend more bucks ;)
Mar 13 10:32:18 <evil-dna> try to install again..
Mar 13 10:32:35 <evil-dna> 16MB seems to be enough i think :/
Mar 13 10:32:45 <evil-dna> besides, perhaps your unit has 32M
Mar 13 10:32:50 <evil-dna> did you check?
Mar 13 10:32:51 <morale> its only 10$ more 
Mar 13 10:32:57 <morale> yeah its only 16mb
Mar 13 10:33:01 <evil-dna> some late v2.2 have 32M with 16M disabled
Mar 13 10:33:10 <evil-dna> hm
Mar 13 10:33:19 <evil-dna> you can change it?
Mar 13 10:33:20 <evil-dna> blah
Mar 13 10:33:35 <morale> nah, if i try i can turn this thing into a brick
Mar 13 10:33:40 <evil-dna> lol
Mar 13 10:33:54 <evil-dna> before you return your unit, make sure to put it back how it was ;)
Mar 13 10:34:00 <evil-dna> hehe
Mar 13 10:34:25 <evil-dna> morale: you enabled boot_wait?
Mar 13 10:34:47 <evil-dna> just be patient and wait for official support
Mar 13 10:34:57 <evil-dna> what if you get a v1.1 gs?
Mar 13 10:35:05 <evil-dna> then you'd need experimental as well ;)
Mar 13 10:35:55 <evil-dna> hello.c:1:19: stdio.h: No such file or directory
Mar 13 10:36:10 <evil-dna> i'm trying to cross compile a hello_world for the wrt54g
Mar 13 10:36:15 <evil-dna> but... :(
Mar 13 10:36:37 <evil-dna> i got a mipsel-ecoff gcc installed tho
Mar 13 10:53:51 <evil-dna> morale: still there,
Mar 13 10:53:52 <evil-dna> ?
Mar 13 11:08:28 <CIA-12> 03jp30 * 10unslung/make/gconv-modules.mk: gconv-modules: install a startup script to ensure that we stay installed over firmware upgrades
Mar 13 12:04:53 <CIA-12> 03jp30 * 10unslung/sources/gconv-modules/control: bump gconv-modules ipk version
Mar 13 12:09:22 <CIA-12> 03jp30 * 10unslung/Makefile: deprecate libiconv
Mar 13 12:37:03 <CIA-12> 03jp30 * 10unslung/Makefile: make BASE_DIR a simply-expanded variable
Mar 13 12:49:58 <CIA-12> 03jp30 * 10unslung/make/ (svn.mk libogg.mk libvorbis.mk): libogg, libvorbis, and svn were all deleting other packages' files from the staging directory. now they don't
Mar 13 13:31:37 <jp30> rwhitby-away, ping?
Mar 13 13:57:35 <dyoung> looks like ctrlproxy kinda works, but I dont have enough flash space to really test it...
Mar 13 14:05:56 <rwhitby-away> jp30: pong
Mar 13 14:11:48 <jp30> rwhitby, bob_tm and I were discussing some changes to template.mk, and I thought would get your comments (and those of anyone else here with views)...
Mar 13 14:12:44 <jp30> ...we thought it might be good to edit template.mk to encourage pkg devs to autogenerate a control file, rather than storing one in sources. sound good?
Mar 13 14:13:55 <rwhitby-away> agreed
Mar 13 14:14:11 <rwhitby-away> did you see TARGET_ARCH?
Mar 13 14:14:42 <dyoung> I think we tossed it around a bit before and everyone said it was a good idea, and your packages already do it.
Mar 13 14:15:29 <rwhitby-away> jp30: did you know that we are reusing the packages as-is for the Asus wl500g router custom firmware? (mipsel)
Mar 13 14:15:54 <jp30> yes, i saw you mention it in the logs and guess that your recent change to template.mk was related to that
Mar 13 14:15:55 <rwhitby-away> 5 line change to top-level Makefile, and 90 packages build
Mar 13 14:16:02 <jp30> nice.
Mar 13 14:16:12 <rwhitby-away> wl500g.dyndns.org
Mar 13 14:16:50 <jp30> i also recently changed BASE_DIR to a simple var (ie. := rather than = ) that speeds up make by about 10x
Mar 13 14:17:08 <jp30> ...very noticable on native
Mar 13 14:17:57 <rwhitby-away> I'm sure there are lots of optimisations waiting in there ...
Mar 13 14:18:52 <jp30> another thing i was thinking about is whether the root Makefile should set and export TMPDIR
Mar 13 14:19:34 <jp30> ...there are some native packages that really thrash swap by filling up /tmp with temporaries
Mar 13 14:22:01 <rwhitby-away> yeah, good idea
Mar 13 14:22:17 <rwhitby-away> a local tmp dir
Mar 13 14:23:42 <rwhitby-away> Oleg (author of wl500g firmware) may well contribute back patches for packages to generate control file (cause he needs to do that to use TARGET_ARCH.
Mar 13 14:23:50 <jp30> so, if I set TMPDIR=$(BASE_DIR)/tmp, add it to the directories target, set it to be exported, and make a rule to create it if it doesn't exist...
Mar 13 14:24:15 <jp30> that would be handy.
Mar 13 14:24:35 <jp30> even my control-less packages have hardwritten arches in them
Mar 13 14:25:36 <rwhitby-away> He has created a basic feed of mipsel ipks since last night: http://wl500g.dyndns.org/unslung/
Mar 13 14:25:55 <rwhitby-away> We're calling it Unslung/WL :-)
Mar 13 14:27:12 <jp30> cool
Mar 13 14:27:25 <jp30> i wonder what else we try porting unslung to
Mar 13 14:27:38 <rwhitby-away> Note this is only a port of the packages, not the firmware
Mar 13 14:27:45 <jp30> ah, of course.
Mar 13 14:27:58 <rwhitby-away> He just sent me a patch for TARGET_ARCH for all the packages (but not generating control files yet)
Mar 13 14:28:58 <jp30> hard working chap, this oleg
Mar 13 14:29:08 <rwhitby-away> and also an ipkg.mk (cause we're only putting the ipkg script in the base wl500g firmware)
Mar 13 14:30:50 <rwhitby-away> so I will check those in now.  he is being careful not to break anything for Unslung/NSLU2.
Mar 13 14:32:04 <jp30> good. this will have been a busy weekend for unslung packages
Mar 13 14:33:29 <rwhitby-away> note that the wl500g is uclibc, so he will probably be contributing patches to make things work with uclibc.  We probably need a TARGET_OS variable so that he can use that to conditionalise patches ...
Mar 13 14:36:10 <jp30> does uclibc have a usable iconv()? if not my recent assault on libiconv may cause trouble
Mar 13 14:36:56 <jp30> ...i'd just got all the unslung packages working with glibc's iconv
Mar 13 14:37:05 <rwhitby-away> dunno - we'll find out.  Note that the needs of Unslung/NSLU2 come first, so if there is any special stuff that needs to be done, then he will do it.
Mar 13 14:43:02 * kergoth groans at the amount of wheel reinvention going on
Mar 13 14:43:08 <Wipmac> hi all. hows the status on the intel-problems with openslug beta release?
Mar 13 14:44:35 <rwhitby-away> kergoth: can OE build stuff for uclibc 0.9.19 ?
Mar 13 14:45:03 <kergoth> if you add uclibc 0.9.19, certainly.  should take, what, 5 minutes?
Mar 13 14:45:08 <rwhitby-away> (I'd love to use OE for this stuff too, but wl500g firmware is like Unslung firmware in that it stays compatible with stock firmware).
Mar 13 14:45:29 <jp30> kergoth, perhaps I will install unslung packages on my zaurus now :)
Mar 13 14:45:38 <rwhitby-away> heh :-)
Mar 13 14:46:05 <kergoth> its your call, i just dont see the point
Mar 13 14:46:39 <jp30> TMPDIR changes to root makefile coming up...
Mar 13 14:47:04 <rwhitby-away> kergoth: I'll get Oleg to look at adding uclibc to OE - but he is right at the bottom of the OE learning curve (starting from zero).
Mar 13 14:47:16 <rwhitby-away> uclibc 0.9.19, that is
Mar 13 14:47:26 <CIA-12> 03jp30 * 10unslung/Makefile: set TMPDIR in root Makefile
Mar 13 14:49:02 <hanjo> hi all. sorry to interrupt you with another of my novice questions: what is the minimum bb set that needs to be rebuilt to get a new kernel-module-blabla.ipk after enabling it in defconfig ?
Mar 13 14:49:30 <rwhitby-away> just rebuild the kernel
Mar 13 14:50:20 <hanjo> rwhitby: I can reproduce the EHCI issue with my char devices on the hub. They work fine
Mar 13 14:51:08 <rwhitby-away> you can or can't?
Mar 13 14:51:20 <hanjo> can't
Mar 13 14:52:19 <rwhitby-away> Dang.  There is a source/ipkg file in the CVS repository (must be in the Attic), which is preventing me adding a source/ipkg directory.  Is there anything I can do other than submitting a sf.net support request and waiting two weeks?
Mar 13 14:57:19 <jp30> it's not possible for a project admin to ssh into sf.net and get access to the repo?
Mar 13 14:57:27 <kergoth> no.
Mar 13 14:57:31 <jp30> hmm
Mar 13 14:57:32 <kergoth> you cant get local access to cvs at sf
Mar 13 14:57:45 <kergoth> afaik anyway
Mar 13 14:58:01 <jp30> that's a pain
Mar 13 14:58:13 <kergoth> yep
Mar 13 15:00:02 <rwhitby-away> jp30: TARGET_ARCH mega-update coming now
Mar 13 15:00:38 <rwhitby-away> Puts @ARCH@ in some control files, and uses sed to change them to ${TARGET_ARCH} - a half-step towards auto-generated control files
Mar 13 15:00:45 <rwhitby-away> also does @VERSION@
Mar 13 15:01:18 <CIA-12> 03rwhitby * 10unslung/ (245 files in 4 dirs): Patched to use TARGET_ARCH - @FOO@ variables put in some control files, and modified on the fly using sed
Mar 13 15:01:22 <CIA-12> 03rwhitby * 10unslung/sources/ (gift/control libid3tag/control): Patched to use TARGET_ARCH - @FOO@ variables put in some control files, and modified on the fly using sed
Mar 13 15:01:34 <CIA-12> 03rwhitby * 10unslung/sources/ (6 files in 6 dirs): Patched to use TARGET_ARCH - @FOO@ variables put in some control files, and modified on the fly using sed
Mar 13 15:01:42 <CIA-12> 03rwhitby * 10unslung/sources/ (openssl/control popt/control samba/control strace/control): Patched to use TARGET_ARCH - @FOO@ variables put in some control files, and modified on the fly using sed
Mar 13 15:01:46 <CIA-12> 03rwhitby * 10unslung/sources/ (tcpdump/control vsftpd/control xinetd/control zlib/control): Patched to use TARGET_ARCH - @FOO@ variables put in some control files, and modified on the fly using sed
Mar 13 15:02:31 <rwhitby-away> jp30: blame me for an errors in that.  I took a quick look at the changes, and Oleg is historically very good, so I haven't recompiled from them or tested.
Mar 13 15:03:20 <jp30> OK, I'll start a big rebuild on the official build machines and see what shakes loose
Mar 13 15:03:36 <rwhitby-away> the changes were trivial, just lots of them
Mar 13 15:03:39 <hanjo> can someone point me to some docs about the build system. I.e. how to invalidate the cache etc. I can't force the kernel to recompile. bb -f openslug-kernel ands with completed without rebuilding
Mar 13 15:04:01 * rwhitby-away points in kergoth's direction :-)
Mar 13 15:04:38 <hanjo> i've checked the OE wiki and the user manual but I am still in the dark
Mar 13 15:04:58 <hanjo> oh, hell I just broke the docs rule :)
Mar 13 15:06:01 <rwhitby-away> kergoth: looks like you've got yourself a volunteer :-)
Mar 13 15:07:00 <hanjo> :)
Mar 13 15:08:32 <hanjo> I just found the work in progress at http://openembedded.org/cgi-bin/moin.cgi/UsersManual
Mar 13 15:08:55 <rwhitby-away> jp30: I;m going to add a TARGET_OS variable (defaults to "linux", but can also take "linux-uclibc") in the top-level Makefile, so that Oleg can use that to switch patches for uclibc.
Mar 13 15:09:29 <jp30> OK
Mar 13 15:10:28 <CIA-12> 03rwhitby * 10unslung/Makefile: Added TARGET_OS variable, so that Unslung/WL500g can use it to conditionally apply uclibc patches.
Mar 13 15:10:39 <rwhitby-away> ok, that's all my changes done.
Mar 13 15:10:45 <rwhitby-away> catch you later.
Mar 13 15:10:56 * rwhitby-away hopes he hasn't broken stuff as he walks away .....
Mar 13 15:11:31 <sam_> So I have a hopefully non-stupid question that wasn't on the faq or the wiki
Mar 13 15:11:58 <jp30> sam_, is it about development, or about how to use unslung?
Mar 13 15:12:22 <sam_> its about how to use it
Mar 13 15:12:30 <jp30> ~nslu2-general
Mar 13 15:12:31 <jbot> #nslu2-general is the IRC channel specifically set up for end-user questions about the NSLU2 device. Questions that are not related to the nslu2-linux development projects should not be asked in #nslu2-linux, but should be asked in #nslu2-general instead.
Mar 13 15:12:34 <sam_> whoops
Mar 13 15:12:41 <sam_> yea.. my bad.
Mar 13 15:12:41 <jp30> s'ok
Mar 13 15:12:50 <sam_> I might be back here some other time ;-)
Mar 13 15:14:39 <rwhitby-away> there
Mar 13 15:15:55 <CIA-12> 03jp30 * 10unslung/make/template.mk: template.mk now encourages developers to generate control
Mar 13 15:17:26 <sam_> perhaps I can phrase my answer in the form a of a development question and thereby skirt the fact that there is no one in nslu-general
Mar 13 15:18:37 <rwhitby-away> that could work :-)
Mar 13 15:19:55 <Gorx> any1 know how to get printer.o to start at boot?
Mar 13 15:21:15 <CIA-12> 03jp30 * 10unslung/sources/xmail/control: fix broken xmail version number in control
Mar 13 15:25:47 <jp30> sam_, you could also post to the yahoo list
Mar 13 15:26:59 <sam_> true that
Mar 13 15:27:38 <sam_> sorry just got distracted answering Gorx's question.
Mar 13 15:27:47 <sam_> irony
Mar 13 15:30:18 <sam_> aha! i just realized i can debug the boot process by setting beeps in the rc.1 script.
Mar 13 15:35:10 <sam_> So the majority of development on the web site seems to be focused around expanding the feature set onto the disk
Mar 13 15:36:24 <sam_> Is there anyone doing any development around keeping the configuration info on the unit and thereby not having to do anything to the disks other than store data?
Mar 13 15:38:11 <jp30> sam_, how about making /opt/etc a symlink to /opt-etc, would that do what you want?
Mar 13 15:38:53 <sam_> well, i was thinking of a little more restructuring than that.
Mar 13 15:39:45 <sam_> more of replacing USB_Detect, modifying some of the cgi, storing users groups and smb.conf on flash.
Mar 13 15:40:33 <sam_> so that USB_Detect2 just mounted /dev/sd[ab]* to say /share/disk*
Mar 13 15:40:40 <jp30> if you really want to modify the firmware in ways that would replace linksys functionality, you probably want openslug
Mar 13 15:40:48 <sam_> hmmm
Mar 13 15:41:48 <sam_> maybe.
Mar 13 15:42:27 <sam_> is there a known issue with a race condition on bootup that might result in a nslu2 not booting fully when there is a disk attached at disk1?
Mar 13 15:43:17 <jp30> what do you mean by not booting fully?
Mar 13 15:43:45 <sam_> i mean reaching a flashing Ready/Status state, but not being accessable via the network
Mar 13 15:43:55 <sam_> the disk1 light comes on
Mar 13 15:44:30 <jp30> your slug is locking up with ready/status flashing?
Mar 13 15:44:53 <sam_> yes, but only if there is a disk attached to disk1
Mar 13 15:45:22 <sam_> and apparently only if I don't put two Set_Led beep1 commands into rc.1
Mar 13 15:45:42 <sam_> which was why I was thinking race
Mar 13 15:47:52 <jp30> you've got to be careful with Set_Led. some things it can do it doesn't return from. beep1 should be OK, though
Mar 13 15:48:20 <sam_> hmmm came up that time.
Mar 13 15:48:29 <sam_> i hate intermittent failures
Mar 13 16:17:24 <sam_> well, thanks for the info.
Mar 13 16:17:56 <sam_> i'm thinking there might be something wrong with this disk, so its off to windows land to test it.
Mar 13 16:17:58 <sam_> yuck.
Mar 13 16:28:29 <CIA-12> 03jp30 * 10unslung/make/cvs.mk: fix sed hackery to generate control
Mar 13 17:41:06 <dyoung> morning sunshine
Mar 13 17:41:32 <[g2]> hey dyoung !
Mar 13 17:42:05 <dyoung> has anyone else successfully booted a extdisk rootfs?
Mar 13 17:42:19 <dyoung> I just cant make it work.
Mar 13 17:42:33 <[g2]> you mean like .sda1root
Mar 13 17:42:56 <CIA-12> 03shenson * 10unslung/ (4 files in 2 dirs): qpopper POP3 server
Mar 13 17:43:02 <[g2]> or booting directly to the external disk
Mar 13 17:44:15 <dyoung> using the .sda1root construct
Mar 13 17:44:32 <[g2]> I've done that a bunch
Mar 13 17:44:42 <dyoung> With the latest 2.6.11.2 ?
Mar 13 17:44:59 <dyoung> or only with 2.6.9 ?
Mar 13 17:46:03 <[g2]> surely 11 and probably several 11.2's
Mar 13 17:46:08 <[g2]> I'm checking 11.2 now
Mar 13 17:46:26 <[g2]> um actually yes with 11.2
Mar 13 17:46:33 <[g2]> I'm pretty sure
Mar 13 17:46:48 <dyoung> oh.
Mar 13 17:47:08 <[g2]> I've got it all ready to go
Mar 13 17:47:10 <dyoung> It doesnt work for me at all.  
Mar 13 17:48:24 <[g2]> Aborting journal on device sda1.
Mar 13 17:48:24 <[g2]> ext3_abort called.
Mar 13 17:48:24 <[g2]> EXT3-fs error (device sda1): ext3_journal_start_sb: Detected aborted journal
Mar 13 17:48:24 <[g2]> Remounting filesystem read-only
Mar 13 17:48:24 <[g2]> SCSI error : <0 0 0 0> return code = 0x70000
Mar 13 17:48:25 <[g2]> end_request: I/O error, dev sda, sector 610
Mar 13 17:48:26 <[g2]> Buffer I/O error on device sda1, logical block 274
Mar 13 17:48:30 <[g2]> that stuff ?
Mar 13 17:48:54 <dyoung> yeah that stuff
Mar 13 17:48:58 <dyoung> cant find [ blahblahblah
Mar 13 17:49:53 <[g2]> did switchbox change ?
Mar 13 17:49:59 <dyoung> No.
Mar 13 17:50:09 <dyoung> Its been at 4.10 for a long time.
Mar 13 17:50:15 <[g2]> did we lose the autofs ?
Mar 13 17:50:19 <[g2]> for root
Mar 13 17:50:25 <[g2]> that was the issue last time
Mar 13 17:50:32 <dyoung> in the fstab?
Mar 13 17:50:38 <[g2]> yeah for /
Mar 13 17:50:43 <dyoung> No hats still good
Mar 13 17:54:10 <[g2]> ummm are you loading the ehci-driver ?
Mar 13 17:55:29 * [g2] guesses there's a race with loading the ehci-hcd driver that rwhitby changed for for OpenSlug
Mar 13 17:55:57 <[g2]> ehci-hcd used to be compiled in
Mar 13 17:57:06 <dyoung> yeah, thats what I gussed too
Mar 13 17:57:09 <dyoung> but I cant make it fit anymore
Mar 13 18:02:57 <[g2]> I'd try loading the ehci-hcd driver just before the sleep 10
Mar 13 18:14:32 <hanjo> hi all. using openslug with 11.2 I get an empty /proc/bus/usb after booting although the host controlers are loaded. Doing a manual mount of usbdevfs fixes the problem. On first look the /etc/fstab entery looks fine (probably usbdevfs should be renamed with the new usbfs). Anyone else seeng the same?
Mar 13 18:15:34 <[g2]> hanjo, I'm wondering about about the noauto on the usbfs in there
Mar 13 18:16:39 <hanjo> defaults 0 0 is what I have on the desktop
Mar 13 18:22:43 <hanjo> g2: any hints how I can recreate the EHCI issue using a flash drive connected to a 2.0 hub? I've tried both reading and writing, but I can not trigger it
Mar 13 18:23:13 <[g2]> hanjo, you've got to do a lot of reading and writing like 250MB at a shot
Mar 13 18:23:32 <[g2]> dd if=/dev/zero of=zero bs=1M count=250
Mar 13 18:23:36 <[g2]> rm zero
Mar 13 18:23:55 <hanjo> I have only 256MB one at hand, with usefull data on it :)
Mar 13 18:24:10 <hanjo> I'll try a large read
Mar 13 18:26:52 <hanjo> hmmm
Mar 13 18:27:11 <hanjo> what fs did you guys use on the HD/FLASH?
Mar 13 18:29:34 <[g2]> ext3
Mar 13 18:39:41 <hanjo> I just finished a 250 MB read from a vfat flash without triggering it
Mar 13 18:43:19 <hanjo> night all
Mar 13 18:43:39 <dyoung> I wonder why it works for you
Mar 13 18:43:42 <dyoung> sigh
Mar 13 18:43:44 <dyoung> okay night.
Mar 13 18:43:58 <hanjo> I'll try some more tests tomorrow
Mar 13 18:44:12 <hanjo> I have a USB HD at work
Mar 13 18:44:16 <[g2]> nite
Mar 13 18:47:32 <hanjo-zzzz> update: I just hit ls in the dir after that and it did happen!
**** ENDING LOGGING AT Sun Mar 13 23:59:56 2005