**** BEGIN LOGGING AT Mon May 16 23:59:56 2005
May 17 00:14:24 <jf-work> morning
May 17 00:15:46 <ka6sox-away> hi there
May 17 00:16:56 <jacques> well that wasn't a very long session was is ka6sox-away ? or have I lost track of time ?
May 17 00:17:32 <ka6sox-away> jacques, well...its 12:17 and we started at 10pm.
May 17 00:17:51 <jacques> ah OK two hours isn't bad, you guys have self-restraint
May 17 00:18:17 <ka6sox-away> (but we missed some people so we played WC3 instead.
May 17 00:18:30 * jacques fondly recalls LAN parties at his house that went all night and into the next morning.
May 17 00:18:44 <jacques> WC3 ?
May 17 00:19:49 <ka6sox-away> WarCraft III
May 17 00:20:10 <jacques> hmmm, winex ?
May 17 00:20:12 <jacques> wine?
May 17 00:20:22 <ka6sox-away> OSX
May 17 00:20:26 <jacques> oooh
May 17 00:20:28 <jacques> duh
May 17 00:20:47 <ka6sox-away> we are a PPC shop here.
May 17 00:23:17 <jacques> i forget that sometimes
May 17 00:24:04 <ka6sox-away> *real* unix :P
May 17 00:24:47 <jacques> ok don't start that :-)
May 17 00:25:00 <ka6sox-away> I won't
May 17 00:25:14 <dyoung-zzzz> As opposed to faux "posix compliance" ?
May 17 00:26:13 <ka6sox-away> jacques, I didn't say that is was better...just "real" (in the sense that it can be certified by the "open group"
May 17 00:26:18 <ka6sox-away> )
May 17 01:31:05 <CIA-1> 03jeanfabrice * 10unslung/ (5 files in 2 dirs): Add rc.nylon, postinst and prerm, bump IPK_VERSION
May 17 01:41:12 <CIA-1> 03jeanfabrice * 10unslung/Makefile: Demote quagga till it gets fixed
May 17 01:53:27 <jf-work> wondering if it is normal to have only 766KB MemFree (16080KB Buffers)
May 17 01:54:34 <dyoung-zzzz> that doesnt sound unreasonable
May 17 01:54:50 <dyoung-zzzz> oits been running for a while right?
May 17 01:55:40 <jf-work> git this just after reboot
May 17 01:55:46 <jf-work> s/git/got
May 17 01:57:04 <jf-work> wiley gives me 1.5MB MemFree after boot, running php-thttpd, nylon and openldap. I'm just running nylon on sluggy
May 17 02:13:08 <jf-work> Did I told I just named my linux toys at home ? Well, we have wiley (classical), haley (wlhdd) and sluggy
May 17 02:14:54 <dyoung-zzzz> Mine is Sluggie. :-)
May 17 02:18:29 <jf-work> I think I will sell haley since sluggy is able to do its job and buy a new wiley dx to replace haley (haley is also doing AP for downstairs)
May 17 02:18:51 <jf-work> my wife gets crazy....
May 17 02:19:17 <dyoung-zzzz> Thats not good.
May 17 02:19:20 * dyoung-zzzz hides
**** BEGIN LOGGING AT Tue May 17 02:40:15 2005
May 17 05:09:56 <CIA-1> 03jeanfabrice * 10unslung/ (make/ipkg.mk sources/ipkg/rc.unslung): Now installs rc.unslung for wiley
May 17 05:11:30 <dyoung-zzzz> woo
May 17 05:26:18 <ziL> I installed my nslu2 yesterday and I can't say anything else than "wow" it's some sweet hardware and the software works flawless (streaming media to my xbox)
May 17 05:27:34 <ziL> The only hickup is the fact that my password didn't stick when I followed the HowTo. I'm going to fix that tonight though.
May 17 07:11:16 <BioK> hi someone here is good enough to help me ?
May 17 07:11:56 <jf-work> BioK: it depends. What's the problem ?
May 17 07:12:15 <BioK> here the situation: on 192.168.1.77 home i see Disk 2/Flash: Ready, 0MB (0% Free)
May 17 07:12:33 <BioK> in administration / status i see Disk 2/Flash:  Type:  SanDisk Cruzer  
May 17 07:12:33 <BioK> Free Space:  0 MBytes (0 %) 
May 17 07:12:59 <BioK> in advanced / disk i see Disk 2 Status: not installed
May 17 07:13:15 <BioK> i can browse file if i click DISK 2(Public Data)
May 17 07:13:38 <BioK> i also can browse file via network neigboor but i can't format the thing i can't adduser it say
May 17 07:14:11 <BioK> warning: No HDD installed, oparation on user, group and share is diseable
May 17 07:14:30 <BioK> if i try to format it say no disk available
May 17 07:14:32 <BioK> any ideal ?
May 17 07:14:34 <jf-work> what firmware are you using ?
May 17 07:14:43 <BioK> Firmware Version:   V2.3R29   
May 17 07:14:50 <jf-work> original one ?
May 17 07:15:02 <BioK> i want to do unslug but i can't right now it say no space left
May 17 07:15:03 <BioK> yes 
May 17 07:15:07 <jf-work> I mean Linksys one ? not unslung ?
May 17 07:15:15 <BioK> yes the linksys one
May 17 07:17:23 <jf-work> Well, I'm pretty new on nslu2 and this channel is dedicated to unslung firmware. I don't know if this is a bug or know feature. Did you check if your firmware is the last one available ? Did you try booting with/without you flashdisk
May 17 07:18:02 <BioK> yes and i got the same thing
May 17 07:25:02 <caplink811_> which size has the stick?
May 17 07:25:23 <BioK> 128
May 17 07:25:58 <caplink811_> well, we should step to #nslu2-general, that the right channel
May 17 07:26:25 <BioK> ok
May 17 08:23:16 <jf-work> nice to see you jp30 !
May 17 08:24:09 <jp30> hi, jf-work
May 17 08:26:14 <jp30> eno-away: if you see this, you might like to look at sqlite.mk to see if you want to build tcl support for sqlite.
May 17 08:27:16 <CIA-1> 03unslung * 10unslung/make/sqlite.mk: sqlite: disable tcl support
May 17 08:40:32 <jp30> jf-work, i'm just trying make quagga-unpack
May 17 08:43:51 <jp30> looks like it is searching for ipv6 headers by grepping the output of cpp. that will work OK as long as cpp is the toolchain cpp, not the host
May 17 08:45:16 <jf-work> Well, I did not investigate on this point. If you see Louis (cannot spell full nickname) and if you have time, could you discuss a bit with him ?
May 17 08:45:53 <jp30> i think that all the author of quagga needs to do is set CPP=$(TARGET_CROSS)cpp. no probs, i will discuss
May 17 08:46:13 <jf-work> ty. I have to leave for home now.
May 17 08:46:18 <jp30> actually, if that solves the problem, it would be appropriate to set CPP=... in TARGET_CONFIGURE_OPTS
May 17 08:46:56 <jf-work> I was fool when I build it on my own machine cause I have ipv6 header. Nudi doesn't
May 17 08:48:03 <jf-work> fooled
May 17 08:48:59 <jf-work> leaving. back in around 4hrs
May 17 08:49:01 <eno> jp30, saw your sqlite.mk change
May 17 08:49:43 <jp30> hi, eno. yeah, it wasn't building correctly on the official host because it was detecting tcl
May 17 08:49:54 <jp30> i figured you didn't intend it to
May 17 08:50:44 <eno> yeah, it's a good idea to be explicit
May 17 08:51:31 <jp30> yes, that's right. perhaps this should be a BestPractices entry: "disable every feature you didn't intend to enable"
May 17 08:53:30 <eno> jp30, i just added erlang to native ready for testing, please note this is an extremely long compile. ~ 11hrs
May 17 08:54:09 <jp30> oh, i see. i started testing it a while ago and thought i would leave it for a bit before assuming a runaway process
May 17 08:54:54 <jp30> is there anything we could do to speed it up? ie. is it using /tmp inappropriately or something?
May 17 08:56:13 <eno> nothing i can think of at the moment. I asked the erlang mailing list, it's very tricky to do cross compilation. Most of its time is spent doing erlc
May 17 08:56:30 <eno> so even distcc cannot help
May 17 08:57:21 <jp30> ok, well, i'll leave it overnight then. i already expect native builds to take decades
May 17 08:57:35 <eno> the lengthy compilation is caused by it's huge feature set. All the snmp, mnesia db, etc.
May 17 08:58:12 <eno> leave it overnite is really a good idea
May 17 08:58:50 <CIA-1> 03unslung * 10unslung/Makefile: set in TARGET_CONFIGURE_OPTS
May 17 08:59:59 <WsasX> hi there
May 17 09:00:40 <WsasX> I upgraded to 4.20 firmware and now am concerned about /sbin/slingover
May 17 09:01:55 <WsasX> On 3.18 I migrated my /opt dir to /share/hdd/data/opt. So what do I have to to? 
May 17 09:27:07 <CIA-1> 03jp30 * 10unslung/Makefile: promote chrpath, sane-backends, ipkg
May 17 11:36:51 <CIA-1> 03bzhou * 10unslung/make/py-mysql.mk: upstream upgrade from 1.2.1c2 to 1.2.1c3
May 17 12:41:46 * bofh_omega say's hi to all
May 17 12:42:44 <bofh_omega> does anyone where I should tweak something if I get an Error like "Cannot find package kernel-module-usbnet" at building bb unslung-image
May 17 12:43:51 <bofh_omega> I know this worked befor but maybe I have broken something usbnet is needed for USB200M Plug to give me a second network interface (works with my current image)
May 17 12:46:06 <bofh_omega> hey don't run away... I'm not THAT ugly ;)
May 17 12:47:57 <bofh_omega> ^^
May 17 12:56:51 <bofh_omega> hmm, maybe something got lost after the hardlockup and following fsck. Is it possible to sync a bk dir (clone: openembedded exists already) or do I have to download it all again (man of the binary crap is't really usefull)
May 17 12:59:25 * bofh_omega likes talking to himself if he is tired :þ
May 17 13:26:59 <bofh_omega> have a nice day/night/life and lots of fun, bye
May 17 13:55:46 <jeanfabrice> llagendiijk: hi
May 17 13:56:00 <llagendiijk> hi
May 17 13:57:21 <llagendiijk> did you have time to test my patches on your build machine? (sorry if I sound pushy)
May 17 13:57:46 <jeanfabrice> llagendiijk: np
May 17 13:57:47 <jeanfabrice> btw
May 17 13:58:04 <jeanfabrice> llagendiijk: could you discuss with jp30 when you see him about your query for adding a TARGET_INCLUDE in top Makefile
May 17 13:58:41 <jeanfabrice> I tried to see explain him today but 1) my english is bad and 2) you know the issue better than I do
May 17 13:58:57 <llagendiijk> ok, will do
May 17 13:59:43 <jeanfabrice> it seems there're many location for include files on slug toolchain
May 17 14:00:00 <llagendiijk> jeanfabrice, your English is not perfect, but definately NOT bad, but I can indeed explain to him WHY we need this
May 17 14:00:36 <jeanfabrice> and perhaps in wiley toolchain too, I did not check that
May 17 14:01:06 <llagendiijk> This should really be the location of the target OS includes (normally /usr/include) 
May 17 14:01:46 <llagendiijk> jeanfabrice, what are the different unslung versions: unslung, whiley etc? What is the wl-500 stuff based on?
May 17 14:02:29 <jeanfabrice> wiley represent the Broadcom based asus family (WiLey = wl)
May 17 14:02:47 <jeanfabrice> wileyware = oleg's firmware (only to be used there)
May 17 14:03:21 <llagendiijk> ok, got it. 
May 17 14:03:30 <jeanfabrice> unslung and openslung are two firmware developped by #nslu2-linux ppl for Linksys NSLU2
May 17 14:03:54 <llagendiijk> But does this mean then that wiley shares a code base with unslung?
May 17 14:04:11 <jeanfabrice> the share the unslung build system
May 17 14:04:36 <jeanfabrice> on the top Makefile, there is a var : UNSLUNG_TARGET. changing it to wl500g or nslu2 change the toolchain
May 17 14:05:39 <jeanfabrice> Linksys SLU2 is based on Intel  XScale-IXP425
May 17 14:06:28 <jeanfabrice> are you ok for being given cvs write access ?
May 17 14:07:42 <llagendiijk> yes, I have not seen rwhitby around, so I could not yet as him
May 17 14:08:05 <llagendiijk> I was planning to ask for it anyhow
May 17 14:49:31 <jeanfabrice> jbot, seen jp30
May 17 14:49:34 <jbot> jp30 <~josh@jp30.nslu2-linux> was last seen on IRC in channel #nslu2-linux, 5h 52m 14s ago, saying: 'ok, well, i'll leave it overnight then. i already expect native builds to take decades'.
May 17 14:59:26 <jeanfabrice> g'night
May 17 15:06:32 <[g2]> jf-asleep, actually it's based on an IXP420
May 17 15:09:50 <jf-asleep> [g2]: /proc/cpuinfo is wrong ?
May 17 15:10:07 <jf-asleep> I'm not very used with slug :)
May 17 15:10:14 <[g2]> I think so
May 17 15:10:24 <jf-asleep> ok
May 17 15:10:44 <jf-asleep> I took the info from there
May 17 15:13:23 <[g2]> http://www.nslu2-linux.org/wiki/Info/CPUOverview
May 17 15:14:22 <[g2]> jf-asleep, IXP420BL is what my chips says
May 17 15:15:18 <[g2]> The RV042 and RV082 Linksys VPN routers are actually IXP425s  one at 266Mhz the other @ 533Mhz
May 17 15:23:59 <NAiL> [g2]: Anyone working with th eRV082 btw? It looks rather nice :)
May 17 15:27:08 <[g2]> NAiL, I've got a 042 and 082
May 17 15:27:24 <[g2]> with some luck I'll be reflashing real soon
May 17 15:27:47 <[g2]> I was running an old version of switchbox 4.11 iirc on the 042
May 17 15:28:52 <[g2]> after telenting in, I think I just used wget for a switchbox.gz unzipped, loopmounted and chrooted in
May 17 15:29:58 <[g2]> NAiL, the boxes are in re-work as of today for JTAG, Serial, MiniPCI on the 042 and extra memory
May 17 15:52:42 <NAiL> [g2]: nice :)
May 17 15:54:21 <CIA-1> 03bzhou * 10unslung/make/samba.mk: explicitly set IFACE_IFCONF to yes for cross build, bumped up IPK_VERSION
May 17 22:12:42 <jbowler-away> If you bk pull any time after now (according to the logs 2005 May 17 20:00 - I don't know if that's UCT) you will also have to pull (svn update) bitbake and delete your bitbake cache (tmp/cache).
May 17 22:13:32 <jbowler-away> The guys on #oe will probably want to know if something doesn't build any more...
May 17 22:30:24 <ka6sox> jbowler-away, what happen?
May 17 22:31:01 <jbowler-away> mickeyl pulled in support for the latest bitbake changes
May 17 22:31:34 <ka6sox> ah.
May 17 22:31:35 <jbowler-away> I believe, but I don't know for sure, that this is actually Holger's fix-the-bitbake-copy problem.
May 17 22:31:46 <jbowler-away> Hum, I can find out easily...
May 17 22:31:47 <ka6sox> k
May 17 22:33:40 <jbowler-away> YES!  It
May 17 22:34:24 <jbowler-away> 42MByte virtual size and it's already parsed 500 .bb files
May 17 22:35:21 <jbowler-away> It isn't any noticeably slower either (it isn't faster of course - speed is limited by parse speed)
May 17 22:36:49 <jbowler-away> May nslu2-kernel .bb files have the setVar problem, that's easy to fix.
May 17 22:36:54 <jbowler-away> s/May/My/
May 17 22:40:49 <jbowler-away> 44.5MByte and 7 minutes to parse 2336 files, building the provider hash is substantially slower, this is expected.
May 17 22:41:55 <jbowler-away> The previous parsing was actually significantly slower I believe - I never timed it.  The previous mem usage was about 600MByte.
May 17 22:46:00 <ka6sox> only 44.5mb! that should make building native *almost* doable!
May 17 22:47:10 <jbowler-away> It should have been before - the 600MByte was just filled with copies of the global metadata, only one copy was required at once so the swapping was minimal.
May 17 22:47:46 <ka6sox> ah
May 17 22:48:02 <jbowler-away> It might get worse now, the data isn't copied so a given bb file will tend to reference more fragmented data.
May 17 22:48:24 <jbowler-away> Still I'm sure the parsing has sped up.
May 17 22:48:25 <ka6sox> so could be slower.
May 17 22:49:44 <jbowler-away> If you want to do native use 'freeze' - see http://www.nslu2-linux.org/wiki/OpenEmbedded/MinimisingBitBakeMemoryUtilisation  You can generate the correct frozen-bbfiles.conf on a non-native system and just copy it over.
May 17 22:50:33 <jbowler-away> It doesn't change much from run to run - it only changes when a new .bb file is added to a directory which is actually used, or when a PREFERRED_VERSION is changed.
May 17 22:52:25 <jbowler-away> So at present openslug-init requires exactly 118 .bb files.  If bitbake just has to handle those it can parse and complete (if nothing needs to be built) in 8s on my system, as opposed to 400s for everything.
May 17 22:53:48 <jbowler-away> Still, the dynamic process size, even with that highly reduced .bb set, is still 40MByte, so it may not help as much as I hoped.
May 17 22:54:32 <jbowler-away> RSS is 6Mbyte - that's promising.
May 17 22:57:13 <ka6sox> yeah..the problem is that even a FAT slug doesn't have enough memory (only 64MB)
May 17 22:57:32 <ka6sox> and that would only run on OpenSlug.
May 17 23:01:54 <jbowler-away> It should be no problem.  Just give it 128MByte of swap.  It isn't a typical program - normally something which is using xMbyte will grind to a halt with less than x/2MByte of physical RAM, but bitbake is really database like.
May 17 23:03:01 <jbowler-away> I.e. it had enormous locality of reference, at least in the original implementation.  The 6MByte RSS indicates that it was running fine on my system with only 6MByte swapped in.
May 17 23:03:03 <ka6sox> okay that is dooable.
May 17 23:04:44 <jbowler-away> There's some stuff in the build which uses perl.  That will probably run out of memory...
May 17 23:22:11 <ka6sox> over 50 downloads of OpenSlug and I haven't seen any feedback on it...
May 17 23:31:22 <jbowler> The "Openslug reboot problem" email thread was about 1.12, but that's all I've noticed.
May 17 23:44:18 <ka6sox> yeah..saw that one...thanks and good night.
**** ENDING LOGGING AT Tue May 17 23:59:56 2005