**** BEGIN LOGGING AT Mon May 09 23:59:56 2005
May 10 00:05:45 <jf-work> 'morning
May 10 00:21:59 <jf-work> jbot, seen rwhitby-away
May 10 00:22:01 <jbot> rwhitby-away is currently on #nslu2-linux.  Has said a total of 22 messages.  Is idling for 18h 14m 32s
May 10 00:32:58 <jf-work> waht exactly is update-alternatives ? I didn't find it in any packages nor in wileyware
May 10 00:39:47 <dyoung> its a ipkg thing
May 10 00:40:02 <dyoung> so when you u pdate something it can clobber or not stuff.
May 10 00:41:37 <jf-work> I'm ipkg v 0.99.145 on wiley
May 10 00:45:15 <dyoung> i dunno too much about it.
May 10 00:45:19 <dyoung> gotta RTFM.
May 10 00:45:47 <jf-work> k
May 10 00:57:01 <jf-work> and is there any good reason why ipkg is not itself an optware ?
May 10 01:11:47 <dyoung> ipkg is maintained by someone else.
May 10 01:12:02 <dyoung> I dont know who offhand but can find out if you like.
May 10 01:12:42 <jf-work> dyoung: I found source on handhelds.org. source are gpl. I guess we can have a package, no ?
May 10 01:13:49 <dyoung> its part of the unslung image.  is Oleg includeing it in Wiley Firmware?
May 10 01:13:55 <jf-work> or I missed something in my young opensource experience
May 10 01:14:01 <dyoung> I should say we include it in the unslung fw...
May 10 01:14:14 <jf-work> Oleg include an ipkg.sh which install the binary
May 10 01:14:20 <dyoung> hehe you would need ipkg to ipkg install ipkg eh.
May 10 01:15:18 <jf-work> yep, but what about new release as update-alternative seems to be present in .148 and not in .145
May 10 01:15:53 <dyoung> Ahh I get what you are asking.
May 10 01:16:00 <dyoung> sorry it was my confusion.
May 10 01:16:18 <jf-work> dyoung: that's me with my engrish, np. I'M confused
May 10 01:16:31 <jf-work> cofee time
May 10 01:22:51 <jf-work> ok. build ipkg package. it produces an ipkg-cl binary.
May 10 01:29:09 <dyoung> that s the ipkg client.
May 10 01:29:24 <dyoung> I'm a little fuzzy on how that all works, ask rwhitby, he'll have the answer.
May 10 01:29:36 <rwhitby> we need to get oleg to include update-alternatives in the ipkg package that he releases
May 10 01:29:43 <rwhitby> he builds that outside of Optware
May 10 01:29:57 <dyoung> Speak of the Devil.
May 10 01:30:03 <dyoung> Good Plan.
May 10 01:30:25 <phedny> does ipkg on wl500g only work with usb-disk attached?
May 10 01:30:32 <rwhitby> then people just go "ipkg upgrade" and they will get the latest ipkg which includes it
May 10 01:30:49 <rwhitby> phedny: you can mount /opt from anywhere (including ramdisk)
May 10 01:31:03 <jf-work> rwhitby: lemme see if ipkg.sh install ipkg does the trick
May 10 01:32:59 <phedny> rwhitby, so in theory i could make ramdisk, perform ipkg operations, store data into mtd partition and load that on boot?
May 10 01:33:09 <jf-work> http://wl500g.dyndns.org/ipkg/ still have the .145 release. Olegs install it from here
May 10 01:33:32 <rwhitby> phedny: yep, if you have the space.  it would have to be a pretty small package
May 10 01:35:45 <rwhitby> can anyone report what /proc/version says on a recent Unslung (a week or so) or OpenSlug (a couple of days) ?
May 10 01:39:26 <jf-work> rwhitby: there is no ipkg optware (for now). How can 'ipkg upgrade'  upgrade ipkg so ? 
May 10 01:39:35 <jf-work> (I can create the ipkg package)
May 10 01:39:48 <rwhitby> cause it gets it from oleg's feed
May 10 01:39:59 <rwhitby> wiley has two feeds, Oleg's base feed, and Optware
May 10 01:40:48 <NAiL> jbowler-zzz: Gotten home. slug still running. Gotta be ntpd ;)
May 10 01:41:45 <jf-work> rwhitby: if we build last optware ipkg, it will take over Oleg's feed
May 10 01:44:44 <james> hello everyone, just wondering if there is any other way to get bitbake as svn is still down..
May 10 01:46:44 <james> has anyone experienced this and has some pointers?
May 10 01:48:18 <dyoung> rwhitby, I flashed with 4.19 dl version.  Is that string h elpful to you?
May 10 01:49:29 <james> can anyone give me some help??
May 10 01:50:34 <jf-work> the ipkg package from Oleg's repo as 'Pierluigi Frullani' as maintainer. Google link this name to handhelds.org so I don't think it was him who create the package
May 10 01:51:02 <rwhitby> dyoung: yep
May 10 01:51:03 <dyoung> james, it tells me the same thing.
May 10 01:51:14 <dyoung> rwhitby, Linux version 2.4.22-xfs (unslung@www) (gcc version 3.4.3) #1 Tue May 3 07:25:48 PDT 2005
May 10 01:51:23 <rwhitby> jf-work: Oleg built it from handhelds.org source
May 10 01:51:33 <rwhitby> with no changes, so he didn't change the maintainer
May 10 01:51:54 <dyoung> jf-work was saying the .148 version has the update-alternatives but the .145 that Oleg uses does not.
May 10 01:52:03 <rwhitby> dyoung: thx - I want to make the @www to be @unslung.org
May 10 01:52:04 <jf-work> so question is : do we create an optware for it ?
May 10 01:52:14 <james> everyone: james needs help so please reply
May 10 01:52:30 <rwhitby> jf-work: just ask Oleg to update to .148 version and include update-alternatives
May 10 01:52:45 <james> everyone: james needs help with getting bitbake, please help him
May 10 01:53:27 <jf-work> rwhitby: ok
May 10 01:53:52 <james> everyone: yeh, that sounds like a good idea! lets all help him with getting svn to work..
May 10 01:55:43 <rwhitby> james: nice :-)
May 10 01:55:54 <rwhitby> for your humour, you are rewarded with this:
May 10 01:56:05 <rwhitby> http://nslu.sf.net/downloads/bitbake.tar.gz
May 10 01:56:34 <rwhitby> it will stay up there only as long as svn is down
May 10 01:56:49 <rwhitby> dyoung: are you able to get an OpenSlug /proc/version too ?
May 10 01:57:54 <james> rwhitby: thanks, now just got to work out what to do with it.... hahaha, any pointers or is that pushing the friendship.
May 10 01:58:07 <james> ahhh, so do i just copy this to my /home/slug?
May 10 01:58:17 <rwhitby> yes, and untar it
May 10 01:59:31 <dyoung> Um.. lemme flash up that image I made last nght.
May 10 02:01:47 <dyoung> Um... I seem to have lost my devslug....
May 10 02:02:01 <dyoung> it might be up to an hour....
May 10 02:04:43 <rwhitby> jf-work: I sent Oleg a message saying you will talk with him about updating ipkg
May 10 02:07:34 <jf-work> [11:01] <jf-work> well, I asked Oleg if he can upgrade ipkg on his feed and propose him to make an optware for it if he doesn't have time to do so.
May 10 02:07:43 <jf-work> (for log)
May 10 02:15:42 <rwhitby> I would prefer that it stay in his feed.
May 10 02:18:28 <dyoung> # cat /proc/version
May 10 02:18:28 <dyoung> Linux version 2.6.11.2 (openslug@downloads) (gcc version 3.4.3) #1 Sun May 8 23:17:03 HST 2005
May 10 02:19:02 <dyoung> I built that last night my time.
May 10 02:27:47 <CIA-1> 03jeanfabrice * 10unslung/ (make/adduser.mk sources/adduser/defconfig): Added su applet to adduser optware
May 10 02:43:52 <rwhitby> dyoung: ok, I need to fix that patch too then
May 10 02:58:04 <jf-work> Is there a quick tip to have "Packages" file regenerated and uploaded to the feed as adduser builds fine, gets uploaded to the feed but Packages still reference the old ipk version
May 10 02:59:04 <james> hello again!
May 10 02:59:12 <james> how do you run a shell script?
May 10 03:07:26 <dyoung> uh....
May 10 03:07:35 <dyoung> care to qualify that question?
May 10 03:12:03 <james> i am trying to run bbdev.sh..
May 10 03:12:09 <james> and i dont know how
May 10 03:12:22 <james> can you impart with me some widsom?
May 10 03:13:52 <rwhitby> james: you might want to read an introductory text on Linux before asking here ...
May 10 03:14:52 <rwhitby> no offence, but if we start answering basic linux questions here, then we won't have time to do anything else.
May 10 03:15:07 <james> rwhitby: okay, thanks for soft landing.. hahaha, i will be back!
May 10 03:15:09 <james> seeya
May 10 03:15:16 <[TEHb]> Hallo Guys!
May 10 03:15:55 <rwhitby> ok, everyone, last chance for Unslung and OpenSlug changes before I tag the release candidate.
May 10 03:17:59 <[TEHb]> Guys, i reading before in ARM linux kernel Mailing List about problem in NSLU2 with bad IRQ handling
May 10 03:18:14 <rwhitby> yes
May 10 03:18:25 <rwhitby> got any insight?
May 10 03:18:42 <[TEHb]> Yeah, i has that problem. I think i can fix it
May 10 03:18:54 <[TEHb]> Or already fixed by anyone?
May 10 03:19:26 <rwhitby> nope, if you're referring to the "irq26: nobody cared" problem, then we haven't cracked it yet and would welcome your input
May 10 03:20:45 <[TEHb]> hmmm, i think i mean other problem. I mean problem, when under heavy ethernet loads, many ethernet frames dropped.
May 10 03:21:16 <rwhitby> R29 reduced the interrupt rate from 4000 to 2000 supposedly to fix that.
May 10 03:21:24 <rwhitby> But we haven't included or tested that yet.
May 10 03:21:53 <[TEHb]> When packets small, and we use full bandwidth, then 50-70% of frames just dropped
May 10 03:22:56 <[TEHb]> R29 ? What is it? I am IXDP425 board developers.
May 10 03:23:14 <rwhitby> ah - Linksys's R29 version of their firmware
May 10 03:23:31 <CIA-1> 03rwhitby 07OpenSlug-1.10-betaUnslung-4.20-beta * r1.3251 10openembedded/ (6 files in 5 dirs): Unslung 4.20-beta release candidate, OpenSlug 1.10-beta release candidate: updated anonymiser patches, and bumped the versions.
May 10 03:24:26 <[TEHb]> rwhitby: R29 fully fix problem with ethernet under heavy load?
May 10 03:24:38 <rwhitby> [TEHb]: Dunno.  I haven't tried it.
May 10 03:24:49 <rwhitby> I personally haven't experienced the problem.
May 10 03:25:43 <[TEHb]> rwhitby: hmmm... i think all use Linksys board for networking purposes :-/
May 10 03:26:13 <rwhitby> others here may well have experienced it
May 10 03:26:21 <rwhitby> just not me :-)
May 10 03:26:37 <rwhitby> what is the fix?
May 10 03:27:53 <[TEHb]> replace APIC on board, i think problem with APIC.
May 10 03:29:56 <dyoung> I have no outstanding changes.
May 10 03:31:18 <dyoung> [TEHb], what do you mean replace the APIC?
May 10 03:31:59 <[TEHb]> dyoung: replace the APIC chip, that can't handle heavy IRQ raising from devices
May 10 03:32:41 <[TEHb]> dyoung: you know other things, how it can be fixed?
May 10 03:34:34 <dyoung> ...
May 10 03:35:12 <dyoung> are you speaking of the triggering? 
May 10 03:35:48 <[TEHb]> dyoung: yeah, triggering
May 10 03:36:02 <dyoung> we've been talking about fooling with that.
May 10 03:36:08 <dyoung> nobody's actualyl done it yet.
May 10 03:37:04 <dyoung> we're running the NPE in polled mode too, btw.
May 10 03:37:58 <[TEHb]> dyoung: then, even polling doesn't help ?
May 10 03:39:03 <dyoung> it helps.  But our main problem is this irq26: nobody cared problem.
May 10 03:39:19 <dyoung> What kernel version are you running?
May 10 03:40:24 <[TEHb]> 2.4.20mvl31, 2.4.27-uc1, 2.6.11.6, 2.6.11.7
May 10 03:40:54 <[TEHb]> and packets dropped on all that kernel not only by NPE
May 10 03:41:10 <dyoung> I can give you the polling patch if you want to compare.
May 10 03:41:15 <[TEHb]> But also i did try use Intel PCI 82559. 
May 10 03:41:55 <[TEHb]> patch for what version?
May 10 03:42:00 <dyoung> 1.4
May 10 03:42:18 <dyoung> um 1.4, but its been patched to hell for 2.6 compatibility.
May 10 03:42:21 <[TEHb]> hmmm, i use 1.5, but i port it to myself
May 10 03:42:40 <dyoung> wait, you have a patch to make 1.5 compile under 2.6?
May 10 03:42:54 <[TEHb]> yeah, i port 1.5 to 2.6 kernel
May 10 03:43:09 <dyoung> wow, can I have that patch?
May 10 03:43:37 <dyoung> I need that for bootloader development.
May 10 03:44:20 <[TEHb]> yep, but i need test it. But today i think i upload it.
May 10 03:44:40 <dyoung> Cool, where will you upload it to?  So I can check tommrow.  Its kind of late here now.
May 10 03:45:18 <[TEHb]> i can email you address
May 10 03:45:49 <[TEHb]> because i need find anonymous server, because Intel License doesn't allow me to share it.
May 10 03:46:14 <[TEHb]> And i do not want to share it on my server :) You understand me? 
May 10 03:46:34 <[TEHb]> Or i can mail it to you, patch is not too big
May 10 03:46:39 <dyoung> Okay
May 10 03:46:47 <dyoung> dyoung8888@yahoo.com
May 10 03:46:53 <dyoung> the polling patch is here:
May 10 03:46:55 <dyoung> http://nslu2-linux.bkbits.net:8080/openembedded/anno/packages/ixp425-eth/files/intdriven.patch@1.1?nav=index.html|src/.|src/packages|src/packages/ixp425-eth|src/packages/ixp425-eth/files
May 10 03:47:43 <dyoung> /intdriven.pat
May 10 03:47:46 <dyoung> ch
May 10 03:48:06 <dyoung> all the patches we've come up with are there
May 10 03:48:16 <dyoung> theres one that enables the use of netconsole too
May 10 03:48:23 <dyoung> I think....
May 10 03:48:32 <dyoung> dunno if that one got pushed or not.
May 10 03:49:24 <NAiL> dyoung: netconsole is pushed, IIRC. I was supposed to sync a day or two ago to get it
May 10 03:49:41 <NAiL> (to try and figure out the x1250 issue)
May 10 03:49:49 <dyoung> I have half a patch written to make 1.5 compile under 2.6 but I never finished.
May 10 03:51:41 <CIA-1> 03rwhitby * r22 10/openembedded/ (25 files in 17 dirs): Synch SVN repo to contents of BK repo for Unslung-4.20-beta and OpenSlug-1.10-beta.
May 10 03:53:45 <dyoung> rwhitby, so r22 is the Real Release Candidate?
May 10 03:54:28 <dyoung> NAiL, hehe, so do you want custody of SlugBug 141?
May 10 03:54:46 <NAiL> dyoung: no thanks. I'm no programmer :(
May 10 03:54:58 <dyoung> well, if you discover anything lemme know since its assigned to me.
May 10 03:54:58 <NAiL> I'd love to be able to though
May 10 03:55:25 <NAiL> dyoung: Yeah. Jbowler and I are comparing notes atm :P
May 10 03:55:45 <dyoung> I didnt even write that code.  I just became the defacto custodian of it since I was the last one to do something to it.
May 10 03:56:08 <NAiL> that is always fun
May 10 03:56:22 <dyoung> thats fine with me though.
May 10 03:56:26 <[TEHb]> dyoung: http://nslu2-linux.bkbits.net:8080/openembedded/anno/packages/ixp425-eth/files/intdriven.patch@1.1?nav=index.html|src/.|src/packages|src/packages/ixp425-eth|src/packages/ixp425-eth/files/intdriven.patch
May 10 03:56:41 <NAiL> maaah
May 10 03:56:54 <[TEHb]> dyoung: this patch for ixp425_eth ver 1.1
May 10 03:57:16 <dyoung> I would like to eventually get it to the point where we can get it into the mainline kernel.
May 10 03:57:23 <NAiL> I've become so used to vi after hacking the slug, I start vi instead of my default editor on my dev-box all the time 
May 10 03:57:24 <[TEHb]> i did port 1.2 to 2.6 too
May 10 03:57:27 <dyoung> [TEHb], that patch is against 1.4.
May 10 03:57:45 <[TEHb]> 2005/03/05 g2 | --- ixp425-eth-1.1-r0/ixp425_eth.c.patched2005-03-05 10:14:39.841274016 -0500
May 10 03:57:48 <[TEHb]> 2005/03/05 g2 | +++ ixp425-eth-1.1-r0/ixp425_eth.c2005-03-05 10:16:13.737999544 -0500
May 10 03:58:10 <[TEHb]> then, for ethernet driver, not for Access Library 1.4
May 10 03:58:17 <dyoung> Thats correct.
May 10 03:58:21 <dyoung> 1.1 goes with 1.4
May 10 03:58:39 <[TEHb]> yeah, but easy working with 1.5, with little hacking :)
May 10 03:58:54 <dyoung> just as 1.2 goes with 1.5
May 10 03:59:15 <dyoung> well I'm still interested in seeing your patch.  
May 10 03:59:16 <[TEHb]> yeah, right
May 10 03:59:30 <[g2]> morning
May 10 03:59:34 <dyoung> I'd like to push that one too, but keep getting distracted with other stuff
May 10 03:59:36 <NAiL> morning
May 10 03:59:40 <dyoung> Hey speak of the distractor!
May 10 03:59:59 <[g2]> :)
May 10 04:00:09 <dyoung> [g2], [TEHb] has a patch to make Intel CSR 1.5 compile under Kernel 2.6.
May 10 04:00:31 <dyoung> I'm hoping he'll mail it to me so I can give 'er a whirl.
May 10 04:00:33 <[g2]> cool.  beewoolie had done that too iirc
May 10 04:00:46 <dyoung> beewoolie's patch compiles the library only.
May 10 04:00:47 <[g2]> more choices are good :)
May 10 04:01:07 <dyoung> I have a patch too; but its nly half done to the point of being workable with openslug.
May 10 04:01:17 <[g2]> [TEHb], have you posted to the arm-linux list latey about ixp425 ?
May 10 04:01:37 <[TEHb]> but this work in BE mode only. For working with LE mode, i need patch 2.6 kernel too. Because no LE support for CSR in 2.6 kernel. 
May 10 04:01:59 <[TEHb]> [g2]: no, because Intel License doesn't give me that rights
May 10 04:02:05 <dyoung> Right, we're only interseted in BE here.
May 10 04:02:49 <[g2]> I didn't mean about the code I meant someone had asked about the eepro100 drivers
May 10 04:03:41 <[g2]> because I did see and issue with the eepro 1000 on an ixp422 board
May 10 04:03:42 <[TEHb]> [g2]: yeah, it's me
May 10 04:04:02 <[TEHb]> [g2]: i has problem with packets dropping under heavy ethernet load
May 10 04:04:03 <[g2]> I saw a cacheing issue iirc
May 10 04:04:16 <[g2]> I'll have to check my notes
May 10 04:04:25 <[TEHb]> [g2]: but now i reading you patch
May 10 04:04:39 <[g2]> my patch ?
May 10 04:04:45 <[TEHb]> [g2]: intdriven.patch
May 10 04:04:49 <[TEHb]> [g2]: it is your patch?
May 10 04:04:50 <dyoung> the polling mode patch for ixp425_eth
May 10 04:04:56 <[g2]> ah...
May 10 04:05:04 <dyoung> Siddy gave it to [g2] who pushed it.
May 10 04:05:09 <[g2]> nod.
May 10 04:05:31 <[TEHb]> hmmm, but patch is obfuscate me
May 10 04:06:11 <[TEHb]> patch just comment line where ethAcc_datapath_poll_setup called
May 10 04:06:34 <NAiL> Anyone have any good resources on BitBake? I'm trying to read the manual online, but all I get is "Document Contains No Data"
May 10 04:10:02 * NAiL starts a build of what appears to be OpenSlug 1.10-beta-rc
May 10 04:12:04 <[g2]> [TEHb], it was a page allocation error under load doing an md5sum of a big iso image over nfs on the GigE network
May 10 04:12:16 <[g2]> md5sum: page allocation failure. order:0, mode:0x20
May 10 04:12:17 <[g2]> [<c00544d0>] (__alloc_pages+0x0/0x3a0) from [<c00577f8>]
May 10 04:12:17 <[g2]> (cache_alloc_refill+0x294/0x4ec)
May 10 04:12:17 <[g2]> [<c0057564>] (cache_alloc_refill+0x0/0x4ec) from [<c0057378>]
May 10 04:12:17 <[g2]> (__kmalloc+0x70/0x8c)
May 10 04:12:17 <[g2]> [<c0057308>] (__kmalloc+0x0/0x8c) from [<c0142648>] (alloc_skb+0x40/0xc0)
May 10 04:12:18 <[g2]>  r4 = C0FF61A0
May 10 04:13:42 <[TEHb]> [g2]: ixp425_eth v.1.1 ?
May 10 04:14:15 <[g2]> No.. that a gigabit card in the minipci slot
May 10 04:16:06 <rwhitby> [TEHb]: are you able to send that patch or not?
May 10 04:16:28 <CIA-1> 03rwhitby * r23 10/openembedded/packages/ (25 files in 14 dirs): Added all the files which had been added to the BK repo since the last synch.
May 10 04:16:53 <rwhitby> we are in the middle of a release, and we need to know ASAP whether your patch is a fix for any of our outstanding problems.  So we need the patch, please.
May 10 04:18:06 <rwhitby> dyoung has provided his email address - are you able to send it to him?
May 10 04:23:47 <[TEHb]> rwhitby: Access Library 1.5 patch ?
May 10 04:24:47 <rwhitby> [20:12] dyoung: wait, you have a patch to make 1.5 compile under 2.6?
May 10 04:24:48 <rwhitby> [20:12] [TEHb]: yeah, i port 1.5 to 2.6 kernel
May 10 04:24:56 <rwhitby> that one :-)
May 10 04:25:16 <[TEHb]> rwhitby: as i say, after testing i share it
May 10 04:26:34 <rwhitby> ok, cool.  Usually here we help each other test, but it's your patch so you decide :-)
May 10 04:28:44 <rwhitby> NAiL: if you got it from BK, then yes that's the release candidate
May 10 04:32:04 <NAiL> ok, nice :)
May 10 04:33:36 <rwhitby> just adjusting the SVN repo to make it the same ....
May 10 04:35:59 <CIA-1> 03rwhitby * r24 10/openembedded/packages/meta/task-python-everything_0.9.0.bb: Removed files since last commit
May 10 04:39:56 <rwhitby> I wonder if I can use svn:externals to share the openembedded dir between openslug and unslung.  Any SVN experts in the house?
May 10 04:40:06 <rwhitby> and anyone know how to tag in SVN ?
May 10 04:47:35 <dyoung> rwhitby, I think you can specify the version at te command line.  svn co -r 23 svn:/blah
May 10 04:47:54 <dyoung> or is that not what you meant?
May 10 04:48:35 <rwhitby> dyoung: no, I meant how to record "OpenSlug-1.10-beta" as a symbolic name for r23.
May 10 04:48:47 <rwhitby> BTW, svn:externals works perfectly :-)
May 10 04:50:21 <CIA-1> 03rwhitby * r25 10/.svnignore: Removed bitbake.
May 10 04:51:28 <CIA-1> 03rwhitby * r26 10/ (/ Makefile): Made bitbake an external
May 10 04:58:02 <dyoung> svn cp trunk ?
May 10 05:02:23 <dyoung> Ah yes. thats pretty much it.
May 10 05:02:34 <dyoung> then when you check it out, you can use -d
May 10 05:03:13 <rwhitby> pardon?
May 10 05:03:22 <rwhitby> oh, right.
May 10 05:03:24 <rwhitby> tags
May 10 05:03:59 <rwhitby> I never really got the hang of that whole "copy the whole tree into a different subdir" style of branching and tagging in svn.
May 10 05:22:54 <rwhitby> I'm gonna call the "tags" directory "releases".  So then the URL looks like svn://svn.berlios.de/openslug/releases/OpenSlug-1.10-beta
May 10 05:26:46 <VoodooZ_Log> Are we ditching bk now?
May 10 05:27:20 <dyoung> tagged releases will be in svn.
May 10 05:27:32 <VoodooZ_Log> i see.
May 10 05:27:34 <dyoung> active dev will follow whichever the #oe folk follow.
May 10 05:27:47 <VoodooZ_Log> and they didn't decide anything yet?
May 10 05:27:53 <dyoung> its not determined yet, but bk is expected to expire in june.
May 10 05:28:20 <dyoung> There was talk of git/cogito, but nothing solid yet afaik.
May 10 05:28:43 <VoodooZ_Log> They'll figure it out.
May 10 05:28:51 <dyoung> Yeah.
May 10 05:30:02 <dyoung> unslung R3 I think.
May 10 05:30:07 <dyoung> unless I'm hallucinating.
May 10 05:30:19 <dyoung> whoops
May 10 05:31:44 <rwhitby> Hmm - I just realised why they have a trunk dir
May 10 05:32:24 <rwhitby> cause "svn copy ..../openslug ...../openslug/releases/OpenSlug-1.10-beta" is just a bit too recursive.
May 10 05:32:46 <rwhitby> So the bad news is that I'm gonna have to change the repo layout again
May 10 05:33:15 <rwhitby> to move the current contents under openslug/trunk/openslug, and then tag the release as openslug/releases/OpenSlug-1.10-beta
May 10 05:34:38 <jacques> ok
May 10 05:38:40 <CIA-1> 03rwhitby * r27 10/ (releases/ trunk/ trunk/openslug/): New layout preparation
May 10 05:40:54 <CIA-1> 03rwhitby * r28 10/ (.svnignore trunk/openslug/.svnignore): Repo layout change
May 10 05:41:05 <CIA-1> 03rwhitby * r29 10/ (Makefile trunk/openslug/Makefile): Repo layout change
May 10 05:41:17 <CIA-1> 03rwhitby * r30 10/ (conf/ trunk/openslug/conf/): Repo layout change
May 10 05:41:27 <CIA-1> 03rwhitby * r31 10/ (openembedded/ trunk/openslug/openembedded/): Repo layout change
May 10 05:41:41 <CIA-1> 03rwhitby * r32 10/ (nslu2-linux/ trunk/openslug/nslu2-linux/): Repo layout change
May 10 06:20:17 <bob_tm> rwhitby: Any thoughts on the disk mount issue?
May 10 06:22:02 <rwhitby> nope, to busy with the release :-)
May 10 06:23:47 <bob_tm> Ok - I have a suggestion, so ping me when ready
May 10 06:24:30 <rwhitby> at this stage, it won't make it into this beta release.  will that be a problem?
May 10 06:25:05 <bob_tm> Nope, we just say that jffs+hd diversion needs tweaking
May 10 06:25:16 <bob_tm> A wiki page will do
May 10 06:25:51 <bob_tm> External root is preferable anyway
May 10 06:28:05 <CIA-1> 03rwhitby * r33 10/trunk/openslug/Makefile: Added additional entries for setup-env for Gentoo users.
May 10 06:28:44 <CIA-1> 03rwhitby * r8 10/trunk/unslung/Makefile: Added additional entries for setup-env for Gentoo users.
May 10 06:52:00 <CIA-1> 03rwhitby * r34 10/trunk/openslug/: Added prop for bitbake
May 10 06:56:23 <rwhitby> if the builds from those complete, and early testing it good, then I'll create the tags tomorrow.
May 10 06:57:26 <CIA-1> 03jeanfabrice * 10unslung/ (Makefile make/usbutils.mk): Promoted usbutils for wiley (ask from Oleg who also give some patches)
May 10 06:57:40 <CIA-1> 03jeanfabrice * 10unslung/sources/usbutils/ (lsusb-wchar-unused.patch control): Promoted usbutils for wiley (ask from Oleg who also give some patches)
May 10 06:57:57 <rwhitby> night all
May 10 06:58:42 <jf-work> 'nite rwhitby-asleep
May 10 07:06:49 <CIA-1> 03bzhou * 10unslung/ (Makefile make/vim.mk): vim ready for cross testing, with enable-gui=no, so no gvim
May 10 07:35:04 <bob_tm> rwhitby-asleep: Made suggestion on http://www.nslu2-linux.org/wiki/HowTo/FlashRootAndDrive
May 10 10:00:30 <Machiavelli> Hi
May 10 10:22:07 <jp30> jf-work, i am making some changes to your changes to usbutils.mk to get it to build for slug
May 10 10:32:07 <CIA-1> 03jp30 * 10unslung/make/usbutils.mk: usbutils: use autoreconf to update ltconfig
May 10 10:36:23 <CIA-1> 03jp30 * 10unslung/Makefile: promote nget py-cherrypy usbutils vim
May 10 12:25:30 <JpNoP> hey
May 10 12:26:09 <JpNoP> mm i have 1 probleme whith an asus 500g deluxe, anyone maybe can help me here ? :/
May 10 12:58:00 <pTweety> can someone please verify the md5sum of unslung-4.19-alpha.bin: 59d22378444dddb16917f09afe8683d1
May 10 12:58:12 <pTweety> I had some troubles with the download
May 10 13:03:56 <bob_tm> pTweety: Download from where?
May 10 13:04:12 <pTweety> http://ipkg.nslu2-linux.org/downloads/
May 10 13:04:57 <bob_tm> pTweety: Just a sec - DL'ing
May 10 13:05:04 <pTweety> thanks
May 10 13:07:11 <jacques> a0efef863a68c8255f1ef71e420d5648  unslung-4.19-alpha.bin
May 10 13:07:33 <pTweety> hm, looks different
May 10 13:07:35 <jacques> oops
May 10 13:07:47 <jacques> stupid POS MF crap galeon browser
May 10 13:07:59 <pTweety> :-))
May 10 13:08:41 <jacques> 59d22378444dddb16917f09afe8683d1  unslung-4.19-alpha.bin
May 10 13:09:07 <pTweety> ah perfect
May 10 13:09:19 <pTweety> but was your download really that fast?
May 10 13:09:29 <pTweety> it took me more than 20 mins
May 10 13:10:46 <bob_tm> Late, but:
May 10 13:10:47 <bob_tm> bash-2.05b$ md5sum -b unslung-4.19-alpha.bin
May 10 13:10:49 <bob_tm> 59d22378444dddb16917f09afe8683d1 *unslung-4.19-alpha.bin
May 10 13:11:05 <pTweety> bob_tm, thank you
May 10 13:11:20 <bob_tm> I got 25KBps on the DL
May 10 13:11:47 <pTweety> hm, than I maybe should blame my isp
May 10 13:12:17 <bob_tm> Yeah - my ADSL line (now in Norway) isn't the best and torrents are running, so I think so.
May 10 13:13:58 <pTweety> now that I sure everything is ok, I'll try to upgrade and put myself into the list of alpha tester :-)
May 10 13:15:34 <jacques> I was getting over 90kB/s
May 10 13:15:52 <jacques> oops
May 10 13:15:54 <jacques> 100%[====================================>] 8,388,608    213.56K/s    ETA 00:00
May 10 13:15:56 <jacques> :-)
May 10 13:18:25 <pTweety> jacques, I already was some kind of jealous. :-))
May 10 13:18:28 <bob_tm> jacques: Hrmf! ;-)
May 10 13:18:44 <ka6sox-office> thats pretty quick!
May 10 13:18:52 <ka6sox-office> (couse that looks like cable)
May 10 13:19:43 <jacques> yeah
May 10 13:19:53 <jacques> I am limited by the remote sites most of the time
May 10 13:20:05 <jacques> I have seen xfer rates well over 600kB/s on close sites
May 10 13:20:28 <jacques> kernel.org can sometimes be very fast
May 10 13:22:16 <NAiL> bob_tm-confused: forvirra? ;)
May 10 13:22:29 <bob_tm-confused> Aha
May 10 13:23:10 <bob_tm> Feel better now.
May 10 13:23:16 <NAiL> good ;)
May 10 13:23:29 <NAiL> where in .no, btw? Oslo?'
May 10 13:24:23 <bob_tm> Yes - Oslo
May 10 13:24:33 <bob_tm> You?
May 10 13:24:52 <NAiL> Trondheim
May 10 13:25:13 <bob_tm> Ah... finally someone from a place with worse weather than I have.
May 10 13:25:34 <NAiL> Hasn't been that much bad weather lately. A lot of sun, actually ;)
May 10 13:26:08 <bob_tm> Hrmf... first Jacques with his DL rate and now this. It is unfair!
May 10 13:26:57 <NAiL> bob_tm: I get ~450k/sec down when I'm not having any torrents up ;)
May 10 13:27:11 <bob_tm> :-(
May 10 13:27:23 <NAiL> You're on NGT as well.. Slow line?
May 10 13:27:34 <bob_tm> ..and the girl is not Skyping me either. Life is hell.
May 10 13:27:42 <bob_tm> Yeah - Charlie 1850
May 10 13:28:04 <NAiL> ah, Delta 3000.. or whatever :)
May 10 13:29:07 <bob_tm> I suddendly feel this urge to make money. It would make my line faster and more attractive.
May 10 13:29:58 <NAiL> yeah
May 10 13:30:16 <NAiL> But then again, whenever you're in a place that doesn't have that fast a line, it really sucks :(
May 10 13:30:29 <NAiL> (Whenever I visit my parents, for instance)
May 10 13:31:37 <bob_tm> Oh yeah - me too. Terrible.
May 10 13:54:16 <pTweety> rwhitby: upgrading from 3.18-beta to 4.19-alpha was working like a charm
May 10 13:54:50 <pTweety> following the README of course ;-)
May 10 13:54:58 <bob_tm> pTweety: Do you have diversion scripts on HD?
May 10 13:55:37 <pTweety> no, not on my develslug
May 10 13:56:15 <bob_tm> pTweety: Ok - 4.19 broke my whole setup because it is not diversion-compatible with Unslung 3
May 10 13:56:54 <pTweety> my slug is a quite new one
May 10 13:57:28 <pTweety> I had just upgraded to 3.18-beta and installed some devel related stuff
May 10 13:58:10 <bob_tm> It's nice - now you have all these device drivers to play with.
May 10 13:59:07 <dyoung> bob_tm, now that you are 4.x enabled, you can try ipkg install kernel-module-videodev aand w3cam and see of all that stuff works.
May 10 13:59:15 <dyoung> I never tested them. :-)
May 10 13:59:49 <bob_tm> Ah, that is a good idea. I will try that at midnight. I have no devslug, but a prodslug that has scheduled maint. at 00:00
May 10 14:00:29 <pTweety> bob_tm: having a first look at it, I agree with you. It's verry nice to look at the result of `free` ;-)
May 10 14:01:26 <pTweety> it seems to be much more
May 10 14:01:41 <bob_tm> dyoung: The two modules and the app (w3cam) are all in the ipkg list. Looking good.
May 10 14:01:50 <bob_tm> pTweety: :-)
May 10 14:02:21 <dyoung> there should be a ov511 in there too and a couple other various camera drivers, but dunno offhand
May 10 14:02:38 <bob_tm> pTweety: Note that hd root should (according to benchmarks) be much faster than flash root. So before root stuff gets in the cache it will be faster now
May 10 14:02:45 <bob_tm> dyoung. Yes - it is
May 10 14:03:07 <bob_tm> dyoung: I only have the crappy AverMedia cam still, which is ov511
May 10 14:03:15 <dyoung> well, lemme know if it works out, or if I need to fix something.
May 10 14:03:48 <dyoung> I only have spca5xx's handy; but I havnt worked out how to include that one into unslung's 2.4 kernel as module.
May 10 14:04:16 <bob_tm> dyoung. That's the one with driver as external code?
May 10 14:04:30 <pTweety> have to go to bed, night all
May 10 14:04:44 <bob_tm> pTweety: 'night
May 10 14:04:48 <dyoung> its not partof the kernel tarball, so the driver code is external.  its not the onewith the binary doo-dad though.
May 10 14:05:09 <bob_tm> dyoung: Then I remember correctly - I compiled it once.
May 10 14:05:59 <dyoung> it compiles.  But it was endian broken until around 2 weeks ago.
May 10 14:06:17 <bob_tm> Ah, that explains why my binary wasn't a big hit :-)
May 10 14:06:34 <jacques> if only someone would endian-fix libusb
May 10 14:06:54 <[g2]> jacques, I think that's been done
May 10 14:07:09 <ka6sox-office> didn't peteru have a stab at that?
May 10 14:07:19 <jacques> I thought he gave up
May 10 14:07:23 <ka6sox-office> ah
May 10 14:07:25 <dyoung> but since then a lot of o ther stuff has been changed; so just need to work out how to make it go.
May 10 14:07:28 <jacques> but I could be wrong
May 10 14:07:34 <dyoung> peteru gave up.
May 10 14:07:39 <ka6sox-office> off to see Dr. Payne, and is associate Dr. Hirt
May 10 14:07:47 <jacques> good luck ka6sox-office 
May 10 14:08:13 <ka6sox-office> Ve Haft 2 Operate!
May 10 14:08:57 <ka6sox-office> oops..wrong channel/.
May 10 14:20:06 <dyoung> odoc!
May 10 14:20:21 <odoc> hi
May 10 14:20:39 <dyoung> Hi.  Did you get my email message from a week or 2 ago?
May 10 14:20:48 <odoc> nope
May 10 14:21:01 <dyoung> Ah that explains a few things.
May 10 14:21:59 <dyoung> what I said was basically VoodooZ was looking for you to ask some quesitons about  your  camera.
May 10 14:22:08 <dyoung> Lemme try send that aagain.
May 10 14:24:12 <odoc> lot of work atm
May 10 14:24:18 <odoc> havent been here for a long time
May 10 14:31:20 <jeanfabrice> 'nite
May 10 14:34:44 <odoc> VoodooZ_work: alive ?
May 10 14:36:02 <jf-asleep> rwhitby: Olegs send me it's source mk & patch for ipkg optware. I've create one but I'm unable to commit it. sf.net return a strange error when running cvs add sources/ipkg
May 10 14:36:16 <jf-asleep> it's = his
May 10 14:45:26 <rwhitby> what error?
May 10 14:46:47 <jf-asleep> cvs commit: failed to create lock directory for `/cvsroot/nslu/unslung/sources/ipkg' (/cvsroot/nslu/unslung/sources/ipkg/#cvs.lock): No such file or directory
May 10 15:22:43 <gerrynjr> http://www.staples.com/Catalog/Browse/Sku.asp?PageType=1&Sku=572863 < nslu2 for 59.99
May 10 15:22:51 <gerrynjr> just figured i'd tell you guys about it
May 10 15:23:02 <gerrynjr> good today only 8(
May 10 15:23:23 <jacques> gerrynjr, wow
May 10 15:23:27 <gerrynjr> and no, i'm not intending to spam here 8-)
May 10 15:23:42 <gerrynjr> I bought mine full price from newegg.com
May 10 15:23:43 <jacques> this is appropriate AFAIAC
May 10 15:23:52 <ka6sox-office> very appropriate.
May 10 15:24:15 <gerrynjr> 59.99 is a steal for these things 8-)
May 10 15:24:42 <gerrynjr> i'd post to the list, but yahoo is being retarded 8-)
May 10 15:24:50 <ka6sox-office> gerrynjr, already GONE
May 10 15:25:00 <ka6sox-office> they didn't last :P
May 10 15:25:06 <jacques> crap
May 10 15:25:11 <gerrynjr> what, it worked here 8-)
May 10 15:25:28 <gerrynjr> gone in your area code maybe
May 10 15:25:30 <jacques> well the price is still good
May 10 15:25:35 <jacques> they are just out of stock
May 10 15:25:44 <jacques> 3-5 days expected availability
May 10 15:26:06 <ByronT> yeah, I think ka6sox and I got some steals of refurb'd ones from newegg a while back
May 10 15:26:06 <gerrynjr> price is good today only though
May 10 15:26:22 <ka6sox-office> yeah...at $70/each
May 10 15:26:26 <jacques> hmm, look at the name in the cart: Linksys NSLU2 Wireless-B Network Storage Link
May 10 15:26:29 <gerrynjr> you may be able to order over the phone
May 10 15:26:32 <jacques> NSLU2 wireless ??
May 10 15:26:35 <ka6sox-office> as a matter of fact I'm putting OpenSlug on one this afternoon.
May 10 15:26:43 <gerrynjr> jacques: it's a typo
May 10 15:27:58 <jacques> yeah I figured
May 10 15:28:15 <gerrynjr> they just went out of stock in my area code
May 10 15:28:20 <rwhitby> They've done with super-secret [g2] wireless mod on them!
May 10 15:28:21 <gerrynjr> unable to order moew
May 10 15:28:26 <gerrynjr> more
May 10 15:28:40 <ka6sox-office> yeah..I just tried as I need 2 more.
May 10 15:28:57 <gerrynjr> ka6sox-away: try other area codes?
May 10 15:29:09 <gerrynjr> you can put in a diff area code and stillg et it shipped right 8-)
May 10 15:29:30 <ka6sox-office> um...didn't know that.
May 10 15:29:37 <gerrynjr> kill the cookies, chnage to a diff one
May 10 15:30:00 <gerrynjr> it's a loophole int he site atm
May 10 15:30:15 <gerrynjr> and my space key is being retarded today
May 10 15:31:06 <ka6sox-office> gerrynjr, still out of stock...I put in boston, la, sandiego, seattle and other placces...no go.
May 10 15:31:11 <jacques> yeah
May 10 15:31:14 <jacques> :-(
May 10 15:31:33 <gerrynjr> ka6sox-office: try calling?
May 10 15:31:40 <gerrynjr> sometimes they can hold an online raincheck
May 10 15:31:52 <gerrynjr> had them do that w/ a deal on a 250 gig harddrive once
May 10 15:33:03 <ka6sox-office> I'm trying right now.
May 10 15:33:06 <ka6sox-office> calling that is.
May 10 15:34:28 <gerrynjr> hopefully all goes well..
May 10 15:34:54 * gerrynjr has still yet to solder on his serial connection
May 10 15:36:58 <ka6sox-office> gerrynjr, called the online center (they are OUT) called the store, they have them BUT its $100
May 10 15:37:18 <gerrynjr> ka6sox-away: store wont pm to the web?
May 10 15:37:36 <gerrynjr> pm == pricematch
May 10 15:37:53 <jacques> nah they never do :-(
May 10 15:38:01 <gerrynjr> *sometimes* they do
May 10 15:38:06 <gerrynjr> ok, now your only hope..
May 10 15:38:18 <gerrynjr> is to find another web retailer.. that sells these things...
May 10 15:38:25 <gerrynjr> and pm to staples
May 10 15:38:37 <gerrynjr> man, this sucks 8(
May 10 15:39:09 <ka6sox-office> yeah.
May 10 15:40:11 <gerrynjr> in any case, i tried to pass the message as soon as I saw it, sorry about that
May 10 15:40:29 <jacques> np, you tried :-)
May 10 15:40:31 <ka6sox-office> gerrynjr, you did well....no problem1
May 10 15:40:49 <jacques> slugs almost never go on sale
May 10 15:41:17 <ka6sox-office> unless they are refurb.
May 10 15:41:51 <gerrynjr> ka6sox-away: curious, got an amex card?
May 10 15:42:04 <gerrynjr> there's one more way 8-)
May 10 15:42:13 <ka6sox-office> I *do* have an amex card.
May 10 15:42:19 <gerrynjr> amex blue?
May 10 15:42:25 <ka6sox-office> Green
May 10 15:42:33 <gerrynjr> 8(
May 10 15:42:35 <gerrynjr> no go then
May 10 15:42:39 <ka6sox-office> k
May 10 15:42:44 <gerrynjr> blue allows for pricematching
May 10 15:42:50 <ka6sox-office> interesting.
May 10 15:47:03 <gerrynjr> if you bought 2 I could have gotten you a coupon for 20 off 100 too 8(
May 10 16:01:57 <rwhitby> svn co svn://svn.berlios.de/openslug/releases/OpenSlug-1.10-beta
May 10 16:02:05 <rwhitby> svn co svn://svn.berlios.de/unslung/releases/Unslung-4.20-beta
May 10 16:02:22 <rwhitby> those are the source snapshots, from which the beta releases will be built.
May 10 16:03:20 <rwhitby> all alpha testers, please verify you can build from those source snapshots.  Should be as easy as "make"
May 10 16:05:41 <jacques> I started an openslug build, I'll start unslung on another box
May 10 16:05:55 <rwhitby> and all previous "pre-release" binary images have been removed from nslu.sf.net/downloads
May 10 16:07:24 <rwhitby> please report status here as you test.  thx.
May 10 16:07:27 <gerrynjr> rwhitby: all the unsling beat images?
May 10 16:07:41 <gerrynjr> beta
May 10 16:08:01 <rwhitby> no, the pre-release IRC-special-secret binaries only
May 10 16:08:14 <gerrynjr> ah 8-)
May 10 16:08:18 <rwhitby> the 1.x, 2.x and 3.x betas are still there on sf.net
May 10 16:08:19 <gerrynjr> wont affect me then
May 10 16:08:23 <rwhitby> (in the file releases area)
May 10 16:08:29 * gerrynjr is still running 2.x
May 10 16:08:38 <gerrynjr> 3.x enabled flash writes... didnt like that
May 10 16:08:50 <ByronT> then you should be happy with 4.x
May 10 16:09:03 <gerrynjr> back to the old-er way?
May 10 16:09:45 <rwhitby> we had to go through 3.x and flash to be able to get to the technology required for 4.x external rootfs with more free RAM
May 10 16:10:01 <gerrynjr> ah, i see 
May 10 16:10:14 <gerrynjr> oooo external root
May 10 16:10:17 <gerrynjr> woohoo 8-)
May 10 16:10:24 <gerrynjr> *very* good news
May 10 16:10:41 <gerrynjr> will leave the door open for other "distros" to surface
May 10 16:11:11 <rwhitby> I think you're thinking of external kernel, not external root
May 10 16:11:17 <gerrynjr> (gentoo,rpm based, etc)
May 10 16:11:23 <gerrynjr> rwhitby: nah.. ecternal root
May 10 16:11:26 <gerrynjr> external
May 10 16:11:37 <rwhitby> but you can't change the kernel without a reflash
May 10 16:11:39 <gerrynjr> change the "root" you can change the underlying os
May 10 16:11:49 <gerrynjr> you could still keep the same kernel
May 10 16:12:02 <gerrynjr> I mean, atm, you could also just chroot
May 10 16:12:08 <rwhitby> ok, cool.  within those constraints, yes.
May 10 16:12:23 <rwhitby> back later - good luck with the testing everyone
May 10 16:17:15 <ka6sox-office> please send reports here or in the ML.
May 10 16:19:08 <jacques> ok unslung release build started too
May 10 16:21:00 <ByronT> Please preface the reports with "RELEASE ISSUE" so that it will be easier to catch them in channel
May 10 16:25:36 <ka6sox-office> nite Bob
May 10 17:18:39 <jacques> real    57m33.605s  unslung svn release build
May 10 17:22:57 <NAiL> Is "openslug 1.10-beta builds just nicely and runs the same a "release issue""? ;)
May 10 17:23:19 <ka6sox-office> um...no..that is called a sucess story.
May 10 17:23:37 <ka6sox-office> please update the database with version and date.
May 10 18:11:15 <james> can anyone tell me why image creation fails when it reaches bison?
May 10 18:11:49 <NAiL> james: Could you be a bit more specific?
May 10 18:12:02 <NAiL> NOTE: package bison-native-2.0-r0: task do_compile: started
May 10 18:12:14 <NAiL> Compiling it as we speak, actually
May 10 18:12:19 <NAiL> done :)
May 10 18:12:35 <james> ahhh, so just re-run?
May 10 18:12:47 <NAiL> Dunno... how did it fail?
May 10 18:13:04 <james> i was just trying to fetch the packages.. and it come up with...
May 10 18:13:09 <james> wait for it..
May 10 18:13:39 <james> NOTE: package bison-native-2.0: started
May 10 18:13:39 <james> NOTE: package bison-native-2.0-r0: task do_fetch: started
May 10 18:13:40 <james> NOTE: fetch ftp://ftp.gnu.org/gnu/bison/bison-2.0.tar.gz
May 10 18:13:40 <james> NOTE: fetch ftp://mirrors.kernel.org/gnu/bison/bison-2.0.tar.gz
May 10 18:13:41 <james> NOTE: fetch ftp://ftp.matrix.com.br/pub/gnu/bison/bison-2.0.tar.gz
May 10 18:13:41 <james> NOTE: fetch ftp://ftp.cs.ubc.ca/mirror2/gnu/bison/bison-2.0.tar.gz
May 10 18:13:42 <james> NOTE: fetch ftp://sunsite.ust.hk/pub/gnu/bison/bison-2.0.tar.gz
May 10 18:13:44 <james> NOTE: fetch ftp://ftp.ayamura.org/pub/gnu/bison/bison-2.0.tar.gz
May 10 18:13:46 <james> NOTE: fetch http://www.oesources.org/source/current/bison-2.0.tar.gz
May 10 18:13:48 <james> NOTE: Task failed: Fetch failed: ftp://ftp.gnu.org/gnu/bison/bison-2.0.tar.gz
May 10 18:13:50 <james> NOTE: package bison-native-2.0-r0: task do_fetch: failed
May 10 18:13:52 <james> ERROR: TaskFailed event exception, aborting
May 10 18:13:54 <james> NOTE: package bison-native-2.0: failed
May 10 18:13:56 <james> ERROR: Build of unslung-image failed
May 10 18:13:58 <james> end of transmission
May 10 18:14:48 <NAiL> I can fetch it just fine from the oesources link
May 10 18:15:14 <NAiL> try fetching it manually and see where it fails
May 10 18:15:37 <james> hmmm, i tried to download it from the above link and it works and then i put it in the sources directory in bitbake but it still fails
May 10 18:15:48 <james> should i reinstall bitbake?
May 10 18:16:49 <rwhitby-web> "sources" directory?  shouldn't that be "downloads" ?
May 10 18:17:20 <james> ummmm...
May 10 18:17:31 <james> don't all the files go to souces
May 10 18:17:36 <james> sources?
May 10 18:18:15 <james> "download a pile of sources, by default into the directory ${OEROOT}/sources."
May 10 18:18:36 <rwhitby-web> where does it say that?
May 10 18:19:02 <rwhitby-web> if you're building from the svn snapshot, then that quote is not correct.
May 10 18:19:03 <james> in the getting started tutorial
May 10 18:19:14 <james> ahhhh
May 10 18:19:17 <james> i see..
May 10 18:19:49 <james> so when i run "bitbake -cfetch unslung-image"
May 10 18:20:06 <james> what happens??
May 10 18:20:24 <rwhitby-web> ok, time out.  we're talking about two different things
May 10 18:20:58 <rwhitby-web> ignore everything I said and most things other people are saying right now if you are following the getting started.  people are focused on testing the release snapshot at the moment
May 10 18:21:15 <james> ok.. 
May 10 18:21:45 <rwhitby-web> Start with an empty directory.  "svn co svn://svn.berlios.de/unslung/releases/Unslung-4.20-beta ; cd OpenSlug-4.20-beta ; make"
May 10 18:22:48 <rwhitby-web> oops
May 10 18:23:00 <james> oops??
May 10 18:23:07 <rwhitby-web> that should be "svn co svn://svn.berlios.de/unslung/releases/Unslung-4.20-beta ; cd Unslung-4.20-beta ; make"
May 10 18:23:07 <james> oops, isnt good..
May 10 18:23:12 <jstueve> cd Unslung-4.20-beta
May 10 18:23:22 <rwhitby-web> right
May 10 18:23:39 <jstueve> going beta on both at the same time... impressive
May 10 18:24:32 <james> so what am i doing exactly?
May 10 18:24:33 <james> hahaha
May 10 18:25:26 <rwhitby-web> Start with an empty directory.  "svn co svn://svn.berlios.de/unslung/releases/Unslung-4.20-beta ; cd Unslung-4.20-beta ; make"
May 10 18:25:45 <james> yeh i have done that, but what is this process doing?
May 10 18:26:25 <james> i know it is "checking out" and "changing a directory"
May 10 18:27:45 <james> sorry, if i am being a time waster but i just want to get involved..
May 10 18:29:06 <jstueve> they've made a make file that handles all the build steps in the Unslung-4.20-beta directory, and that should start the compilation, resulting in a unslung image ready for flashing.
May 10 18:29:40 <james> awww... that isnt very hands on..
May 10 18:30:25 <NAiL> no, but it saves lazy guys like me from doing to much ;)
May 10 18:30:43 <NAiL> too
May 10 18:31:51 <james> i am very lazy, but i want to do it like everyone else and hopefully pick up a bit of know-how along the way!
May 10 18:32:08 <NAiL> Everyone else is doing it this way now ;)
May 10 18:32:15 <james> hahaha, what a dork i am
May 10 18:32:31 <james> yeh, looks like i will be waiting for the next alpha
May 10 18:32:41 <james> or i will try my hand at open
May 10 18:33:00 <jstueve> openslug is the same process too. :)
May 10 18:33:23 <james> just a make file?
May 10 18:33:33 <jstueve> yep.
May 10 18:33:46 <jstueve> well and all the structure that makes the makefile work.
May 10 18:34:21 <james> well i think i will just go sit in the corner then..
May 10 18:35:30 <jstueve> you can try to get an openslug irc package done
May 10 18:36:50 <james> hmmmm, have no clue what is meant by that but i will try and find out and get back to you..
May 10 18:36:52 <jacques> it's probably no more complex than `bitbake irssi`
May 10 18:36:55 <rwhitby-web> james: you can port Unslung to the R29 source code from Linksys if you're looking for something difficult to do ....
May 10 18:37:18 <jacques> difficult and pointless?
May 10 18:37:38 <james> once again hmmm...
May 10 18:37:46 <rwhitby-web> jacques: we don't know what improvements might have been made in the executables for which we have no source ...
May 10 18:38:01 <rwhitby-web> (or what new bugs might have been introduced)
May 10 18:38:32 <rwhitby-web> and we at least need to update the web interface so that it shows that it came from R29 so we don't get endless "why are you using R25 when R29 is out" questions ....
May 10 18:39:26 <james> mmm, so what is involved in porting?
May 10 18:45:27 <rwhitby-web> james: that's the task :-)
May 10 18:46:35 <james> rwhitby-web: but does it just mean looking and differences in the code??
May 10 18:46:44 <james> sorry if i am way off the mark
May 10 18:47:24 <rwhitby-web> it means rebuilding the unslung firmware based on the R29 root filesystem contents instead of the R25 root filesystem contents, and then testing the resulting firmware.
May 10 18:49:19 <james> is there a place where i can read about this?
May 10 18:49:30 <james> or something similiar?
May 10 18:52:23 <dyoung-web> What needs testing?
May 10 18:52:55 <dyoung-web> .. what needs testing more?  Openslug or Unslung?  I can fire one off now.
May 10 18:52:58 <jacques> dyoung-web, unslung 
May 10 18:53:20 <dyoung-web> Okay getting it.
May 10 18:55:05 <jacques> did someone still need /proc/cpuinfo from latest unslung ?
May 10 18:55:25 <dyoung-web> rwhitbvy was interested in/proc/version
May 10 18:55:32 <dyoung-web> to see the build id.
May 10 18:55:53 <dyoung-web> it should be unslung@unslung.org or someting like that.
May 10 18:56:47 <jacques> # cat /proc/version
May 10 18:56:47 <jacques> Linux version 2.4.22-xfs (unslung@unslung.org) (gcc version 3.4.3) #1 Tue May 10 17:05:48 PDT 2005
May 10 18:57:09 <rwhitby-web> that's a winner
May 10 18:57:11 <dyoung-web> thats to verify the anonymiser patch does what its supposed.
May 10 18:57:17 <dyoung-web> (which it does)
May 10 18:57:35 <rwhitby-web> jacques: do you have one from Openslug too?
May 10 18:58:37 <dyoung-web> I see some deprecated symlinks
May 10 18:59:15 <rwhitby-web> dyoung-web: what's that?
May 10 18:59:20 <jacques> rwhitby-web, no, my last flash was right before the anonymizer patch
May 10 18:59:27 <dyoung-web> apache, hotplug, muxsshssl
May 10 18:59:41 <rwhitby-web> ah. ok.  won't affect the release
May 10 18:59:47 <rwhitby-web> just fix them in bk
May 10 19:00:03 <dyoung-web> yeah, I'll get to it later.  Just pointing out now so I dont forget.
May 10 19:00:05 <rwhitby-web> and we'll pick them up on the next beta
May 10 19:01:42 <NAiL> rwhitby-web: anonymizer works on openslug as well
May 10 19:01:57 <NAiL> Can't copy and paste atm, but I checked earlier today
May 10 19:02:01 <jacques> real    170m32.970s openslug svn release build completed
May 10 19:06:02 <rwhitby-web> cool, thx
May 10 19:06:41 <jstueve> failed 
May 10 19:06:50 <NAiL> where this time? ;)
May 10 19:06:52 <jstueve> openslug build
May 10 19:07:05 <jstueve> gcc-cross-3.4.3-r10
May 10 19:07:10 <jstueve> do compile
May 10 19:08:33 <jacques> jstueve, run out of space?
May 10 19:09:12 <jstueve> shouldn't have... but it could be...
May 10 19:10:20 <jstueve> got plenty of space
May 10 19:10:36 <jstueve> 10334596 bytes avail
May 10 19:11:01 <NAiL> That's not exactly plenty when building gcc
May 10 19:11:14 <NAiL> infact, it's by far too little
May 10 19:11:16 <jstueve> I think I foluled up trying to c&p
May 10 19:11:35 <jstueve> 31% used...
May 10 19:11:48 <jstueve> make[4]: *** [wlocale-inst.lo] Error 1
May 10 19:11:48 <jstueve> make[3]: *** [all-recursive] Interrupt
May 10 19:11:48 <jstueve> make[2]: *** [all] Interrupt
May 10 19:11:48 <jstueve> make[1]: *** [all-target-libstdc++-v3] Interrupt
May 10 19:12:12 <rwhitby-web> my guess is missing required software
May 10 19:12:54 <jstueve> I'm thinking I hit ctrl+c
May 10 19:12:57 <jstueve> *bluch*
May 10 19:13:09 <jstueve> s/bluch/blush
May 10 19:14:04 <rwhitby-web> PEBKAC
May 10 19:14:19 <NAiL> hihi
May 10 19:14:39 <jstueve> PEBKAC?
May 10 19:14:51 <ByronT> that's an old telephone tech support's acronym
May 10 19:14:55 <NAiL> Problem Exists Between Keyboard And Chair
May 10 19:14:59 <jstueve> check.
May 10 19:15:03 <jstueve> I knew...
May 10 19:15:54 <jstueve> okay, check off mid-comile interrupt, and restart as good to go.
May 10 19:16:30 <jstueve> ID-10-T error.
May 10 20:08:42 <happ> I have terminal access through a usb/serial device (AddASerialPort). I don't see it mentioned anywhere, but it seems I have to connect via that interface and hit enter once from cold boot to the box to continue into Redboot
May 10 20:13:19 <dyoung-web> Really?
May 10 20:13:28 <dyoung-web> Okay you might need the pullups...
May 10 20:14:18 <dyoung-web> for me, when I press the button, a   apperas instantantly.  The redboot comes around 8 sec later.
May 10 20:14:26 <dyoung-web> a (plus sign)
May 10 20:14:52 <jacques> heh
May 10 20:16:51 <jacques> I also have one slug on a usb->serial console and I don;t have to do that
May 10 20:20:59 <dyoung-web> my unslung 4.20 built.  I was gonna try flash it, but I think devslug is turned off.
May 10 20:27:32 <dyoung-web> nope, it had openslug on it.  flashing now
May 10 20:27:51 <dyoung-web> any ideas on how to enable telnet without a browser?
May 10 20:28:12 <dyoung-web> like wget http://slug/Management/telnet.cgi with options?
May 10 20:29:45 <dyoung-web> oh well, flashfailed.
May 10 20:29:52 <dyoung-web> needs a power cycle.
May 10 20:38:49 <rwhitby-web> dyoung-web: is that a real failure, or a user error?
May 10 20:39:25 <dyoung-web> it failed in upslugging it
May 10 20:39:54 <dyoung-web> Not uncommon for me to see that mode of failure when theres lots of other traffic on the same net.
May 10 20:40:07 <rwhitby-web> cool
May 10 20:40:20 <dyoung-web> someone is slamming the hell out of my ssh port now.
May 10 20:41:14 <dyoung-web> I wonder what bug theyre trying to exploit.
May 10 20:41:34 <rwhitby-web> that's me gettng my pingus back :-)
May 10 20:42:00 <dyoung-web> he he
May 10 20:42:59 <dyoung-web> what are those AVI's saved as anyways?  a MPEG PS encasulated in AVI?
May 10 20:46:01 <rwhitby-web> dunno - whatever topf2ps outputs
May 10 20:46:19 <rwhitby-web> the neuston plays them, and that's all I care about :-)
May 10 20:46:28 <rwhitby-web> actually windows media player played them too
May 10 20:46:28 <dyoung-web> hah.
May 10 20:46:43 <dyoung-web> Amazon wont ship me a the actiontec.
May 10 20:46:54 <jstueve> why?
May 10 20:47:21 <dyoung-web> they claim I'm not in the US so they cant ship outside the country.
May 10 20:47:25 <rwhitby-web> dyoung: skunkworks would ship you a neuston ...
May 10 20:49:56 <rwhitby-web> so expansys is being very slow on my wl530g, and I'm almost to the point at looking at some other toy instead.
May 10 20:50:28 <dyoung-web> the internet just got very very slow for me
May 10 20:51:42 <rwhitby-web> http://www.skunkworks.com.au/product_info.php?cPath=2_3&products_id=31 AUD$382 shipped to you
May 10 20:54:38 <jstueve> actiontec is now on my wishlist.. $149 (USD)
May 10 20:54:47 <happ> dyoung-web: thx for the tip on the pullups, I'll try that later
May 10 20:59:46 <rwhitby-web> dyoung-web: what about someone in the continental US purchasing and shipping to you?
May 10 20:59:51 <rwhitby-web> doh' he's gone.
May 10 21:02:25 <jstueve> <--- is in the cont us
May 10 21:39:37 <jstueve> what does a red status light indicated... it turned red and beeped once.
May 10 21:40:00 <jbowler-zzz> openslug?
May 10 21:40:06 <jstueve> yeah...
May 10 21:40:11 <jstueve> not the latest release... 
May 10 21:40:15 <jbowler-zzz> you be a-booting
May 10 21:40:29 <jbowler-zzz> when it goes green you can party
May 10 21:40:30 <jstueve> was trying to upslug but it couldn't find the slug
May 10 21:40:54 <jbowler-zzz> ?  it reset
May 10 21:41:00 <jstueve> it went green..
May 10 21:41:08 <jstueve> haven't upslugged yet
May 10 21:41:44 <jbowler-zzz> something caused the slug to drop out of upgrade mode and boot the kernel, hum, as it flashing red-green before or was it amber?
May 10 21:42:03 <jstueve> it is green now, pingable.. but I can't see it with upslug...
May 10 21:42:11 <jstueve> either on the lan, or direct cable.
May 10 21:42:45 <jbowler-zzz> that's correct - as expected - you should be able to ssh in, it's running fully (i.e. the single user boot has completed and logins are permitted)
May 10 21:43:10 <jstueve> okay... how do I reflash with the released beta then?
May 10 21:43:23 <jbowler-zzz> you can't upslug.  At moments like these I rip it's heart out, i.e. push the power button.
May 10 21:44:01 <jstueve> did that twice...
May 10 21:44:07 <jbowler-zzz> then you must trap the RedBoot prompt, to be able to 'upgrade'
May 10 21:44:42 <jbowler-zzz> Ah... did it twice... impatient eh?  Try this (UN*X) "until telnet 192.168.0.1 9000; do :; done"
May 10 21:45:37 <jbowler-zzz> Wait till the telnet says "host not reachable" - this may take a minute or so - then restart the slug (i.e. press the power button, or plug the power back in and do same.)
May 10 21:46:07 <jstueve> a little,.. I wanted to flash it before I went to bed...
May 10 21:46:47 <jbowler-zzz> xy chromosome problem
May 10 21:46:50 <jstueve> lol
May 10 21:48:05 <jbowler-zzz> RedBoot supposedly waits three seconds looking for a packet on the ethernet before it boots, but lots of people have problems because of the surrounding network.
May 10 21:48:30 <jstueve> it got an arp response... but didn't trap it.. I'm using the pl script...
May 10 21:49:14 <jstueve> it existed with :"bash: a: command not found...
May 10 21:49:32 <jbowler-zzz> I find if it doesn't work one time I have to wait for 192.168.0.1 to become inaccessible before it will work again.  I think it's a problem with some routers caching the route - one failure and you have to wait for the cache to flush.  I use WRT54G's...
May 10 21:50:15 <jbowler-zzz> ? what exited - sorry, I lost track - the perl script?
May 10 21:50:27 <jstueve> yeah...
May 10 21:50:48 <jstueve> I'll try hard cable again...
May 10 21:50:52 <jbowler-zzz> Try the 'until' rune (assuming you are using bash or a bash-alike)
May 10 21:51:15 <jstueve> bash
May 10 21:51:20 <jstueve> the until rune?
May 10 21:51:46 <jbowler-zzz> until telnet 192.168.0.1 9000; do :; done
May 10 21:51:58 <jbowler-zzz> copy'n'paste into your UN*X shell
May 10 21:55:14 <jbowler-zzz> expect to see (maybe after a while): telnet: Unable to connect to remote host: No route to host
May 10 21:55:52 <jbowler-zzz> power cycle the slug (turn it on, or type 'shutdown -r now' to the slug itself)
May 10 21:56:25 <jbowler-zzz> when the light goes orange hold down the ctrl key and hover over C
May 10 21:56:51 <jbowler-zzz> when the telnet session says: == Executing boot script in 1.900 seconds - enter ^C to abort
May 10 21:56:55 <jbowler-zzz> do so...
May 10 21:56:57 <jstueve> so that pings that post for 9000 times...
May 10 21:57:05 <jstueve> now it is saying 'no route to host
May 10 21:57:25 <jbowler-zzz> Good, that's what you want, it's pinging port 9000 on 192.168.0.1
May 10 21:58:26 <jstueve> I was watching the light not the prompt...
May 10 21:58:28 <jstueve> doh...
May 10 22:01:00 <jbowler-zzz> This is the worst part, where it seems that there will never be a RedBoot prompt again.  At this point you have to shut down the slug, kill the telnet - use ^] then 'quit' to telnet, *wait*, restart telnet, and only restart the slug when you get 'no route to host'
May 10 22:01:30 <jbowler-zzz> 'cause somehow the telnet keeps alive the bogus route to the slug
May 10 22:02:47 <jstueve> okay redboot promt
May 10 22:02:49 <jstueve> pt
May 10 22:03:14 <jbowler-zzz> 'upgrade' (to the prompt), then upslug will find the slug
May 10 22:04:56 <jstueve> thanks.. it found it...
May 10 22:05:21 <jstueve> can I kill the redboot telnet?
May 10 22:06:55 <jbowler-zzz> yes, though for religious reasons I only kill it when I get the first feedback from upslug that it's actually working.  ^] then quit (this causes telnet to exit with 0, so the shell until loop terminates.)
May 10 22:12:33 <jstueve> how do I check the version?
May 10 22:13:05 <jbowler-zzz> which version?  the one reported by upslug or the openslug one?
May 10 22:13:10 <jstueve> openslug
May 10 22:13:15 <jstueve> want to confirm the build
May 10 22:13:25 <jbowler-zzz> cat /proc/version
May 10 22:14:00 <jstueve> confirmed openslug
May 10 22:14:03 <jstueve> woo!
May 10 22:15:11 <jbowler-zzz> is this the first time you've booted with openslug?
May 10 22:15:23 <jstueve> second.
May 10 22:15:34 <jstueve> need to do the turnup...
May 10 22:15:58 <jbowler-zzz> Built from bitbake?
May 10 22:16:08 <jstueve> yep.
May 10 22:16:20 <jstueve> from the beta release 
May 10 22:17:37 <jbowler-zzz> Wrong question, I meant 'bitkeeper' - i.e. is it bitkeeper or the svn snapshot?
May 10 22:17:43 <jstueve> svn snapshot
May 10 22:21:00 <jstueve> once I run turnup I should reboot?
May 10 22:22:22 <jbowler-zzz> Yes, the slug is still running with the root file system in flash.  turnup creates a file in flash (e.g. /.sda1root) which causes the next boot to use the disk instead.
May 10 22:25:23 <jstueve> if I wanted to try to build irissi do I just need to change the BBFILES to point to the openembedded packages.. to see if it compiles?
May 10 22:27:33 <jbowler-zzz> ... I'm not sure what BBFILES is currently set to, just looking
May 10 22:27:45 <jstueve> nslu2-linux 
May 10 22:28:30 <jstueve> will openslut recognize another usb key and automount it?
May 10 22:28:35 <jstueve> will openslug recognize another usb key and automount it?
May 10 22:28:48 <jstueve> whoops slipothetounge...
May 10 22:31:24 <jbowler-zzz> edit releases/OpenSlug-1.10-beta/conf/local.conf  after the BBFILES := line add 'BBFILES += <path to irssi>/*.bb'  that will be something like /home/slug/releases/OpenSlug-1.10-beta/openembedded/packages/irssi/*.bb
May 10 22:32:27 <jbowler-zzz> no, you have to tell it everything.
May 10 22:35:23 <jbowler-zzz> I'm guessing that 'mount /dev/sdb1 /mnt' might work, assuming it is plugged in to the second USB port (not sure about the '1').
May 10 22:35:27 <jstueve> that build is nice, all the ipks build too...
May 10 22:38:05 <jbowler-zzz> I've been building from bitkeeper, but I've just svn up'ed my svn tree and typed 'make' so I'll see how that goes...  Oops, it failed at flex-native.  How surprising ;-)
May 10 22:38:28 <jstueve> I have the flex file... do you want it?
May 10 22:38:49 <jstueve> (its at NAiL's repos too)
May 10 22:40:19 <jbowler-zzz> ;-)  I have it too.  What I'm going to try to work out how to do is to make the build look somewhere else for it...
May 10 22:40:49 <jstueve> I dumped it downloads, clobbered tmp and restarted the make
May 10 22:41:17 <jstueve> wonder if trying to build unslung will have the same problem...
May 10 22:42:27 <jbowler-zzz> The simple recovery in this case is 'rm -rf */flex*' (in the same directory as make), but it doesn't help because the sourceforge site always gives back the 'flex*failed*' file.
May 10 22:42:53 <jstueve> ah... 
May 10 22:43:42 <jstueve> bitchx compiled... need to see if it will work on the slug (it should)
May 10 22:44:43 <jbowler-zzz> Unslung make is creating its own 'downloads' directory.  make a central /home/downloads directory and ln -s /home/downloads <unslung>/downloads and <openslug>/downloads (mv the downloaded files to /home/downloads first!)
May 10 22:45:19 <jstueve> ah... okay.. same packages
May 10 22:48:17 <jbowler-zzz> No, but there is considerable overlap.  The files in the downloads directory should have unique names - so if /home/downloads/foo is created for openslug and unslung needs a foo it will be the same file (maybe ;-)
May 10 22:48:35 <jstueve> well I give it a try...
May 10 22:48:45 <jstueve> how can you scp a file to the slug...
May 10 22:49:32 <jbowler-zzz> Remember to retain the files you've already downloaded (i.e. mv the files out of the openembedded/downloads directory before deleting it).\
May 10 22:50:17 <jbowler-zzz> Yes. scp to root@slug:/tmp is my preferred way of getting ipks on (when diskless)
May 10 22:50:46 <jstueve> scp root@slug:/tmp filename?
May 10 22:53:58 <jbowler-zzz> scp filename root@slug:/tmp
May 10 22:57:26 <jstueve> okay.. last interuption...
May 10 22:57:36 <jstueve> how do you define a local ipk store?
May 10 22:57:55 <jbowler-zzz> edit /etc/ipkg.conf
May 10 22:58:26 <jbowler-zzz> After the 'root' line add 'source openslug file:/foo'
May 10 22:59:08 <jbowler-zzz> Hum, 'src' not 'source'.  If ipkg complains about the file:/foo I just add ////// until it shuts up.
May 10 22:59:55 <jbowler-zzz> If you've scp'ed the whole of tmp/deploy/ipk then you might be sorry 'cause the Packages file isn't up-to-date
May 10 23:00:46 <jstueve> I can redo later...
May 10 23:00:58 <jbowler-zzz> Run (from the staging/bin/i686-linux directory, or whatever host os you used) "ipkg-make-index -p Packages ." in the ipk directory, see what it changed and copy that across.
May 10 23:01:51 <jstueve> well, bitchx compiles and installs and runs fine... irssi failed compile
May 10 23:02:08 <jbowler-zzz> The make index gets run when the root file system is built, but then packages are added afterwards to it needs to be re-run.
May 10 23:02:39 <jstueve> it seemed like that is what the Makefile did...
May 10 23:03:14 <jbowler-zzz> unslung is rendered inflexible too...
May 10 23:06:05 <jstueve> jbowler: thank you for you help... I'm off to bed, kicked off an unslung build... check in the am
May 10 23:07:12 <jbowler-zzz> g'nite
May 10 23:08:27 <ka6sox-away> rats...missed him.
May 10 23:09:04 <jbowler-zzz> who?
May 10 23:09:23 <ka6sox-away> you sir...
May 10 23:09:40 <jbowler-zzz> damn, me and my big mouth
May 10 23:09:42 <ka6sox-away> which version of uclibc are you using?
May 10 23:09:48 <jbowler-zzz> 0.9.27
May 10 23:09:49 <ka6sox-away> (for openslug)
May 10 23:09:53 <jbowler-zzz> same
May 10 23:10:07 <ka6sox-away> okay so its later than the 0.9.19 used for wiley...
May 10 23:10:43 <ka6sox-away> thanks...sleep well!
May 10 23:10:55 <jbowler-zzz> it has patches too, including the one to make stat(2) return non-zero values in dev and rdev
May 10 23:11:18 <jbowler-zzz> oh, I've been sleeping for about an hour.  Got distracted.
May 10 23:11:38 <ka6sox-away> and you are using 2.6.11-mm4 for your kernel?
May 10 23:12:16 <jbowler> That's one of my three builds, haven't booted it for several days.  Currently I build that with glibc (to support NIS)
May 10 23:12:30 <ka6sox-away> k
May 10 23:13:07 <ka6sox-away> have you tried to build the latest openslug snapshot?
May 10 23:13:11 <jbowler> This flex native thing is a problem - it kills both openslug and unslung
May 10 23:13:32 <jbowler> (Yes, I've just tried the snapshots - hence the flex comment.)
May 10 23:13:43 <ka6sox-away> well...I can fix that...we have a stash at nslu.sf.net/downloads
May 10 23:14:48 <jbowler> Yes, but as it stands a naive user fails the build early on, would be nice to have a temporary fix in the makefile.
May 10 23:15:25 <ka6sox-away> I'm going to suggest that somebody fix the flex.mk file for both images.
May 10 23:15:39 <ka6sox-away> (and check it into bk)
May 10 23:15:52 <ka6sox-away> I'm not currently setup to do that (only svn here)
May 10 23:16:43 <jbowler> The flex .bb is pretty much correct - it uses ${SOURCEFORGE_MIRROR}, that defaults to aleron.dl.sourceforge.net
May 10 23:17:14 <ka6sox-away> we need to use our stash instead.
May 10 23:17:24 <ka6sox-away> nslu.sf.net/downloads
May 10 23:17:57 <jbowler> Ok, let me try hacking SOURCEFORGE_MIRROR to that.
May 10 23:18:44 <ka6sox-away> we don't have everything in that stash..only a few things that we can't get from other sources
May 10 23:20:42 <jbowler> I understand, I'm hoping it will fall back to something else when it finds that it can't get the dl from nslu.sf.net
May 10 23:21:17 <ka6sox-away> k
May 10 23:21:44 <rwhitby-web> we could get the makefile to download from nslu.sf.net directly into the downloads dir.
May 10 23:21:51 <rwhitby-web> before it starts bitbake
May 10 23:21:59 <ka6sox-away> that is a good idea.
May 10 23:22:24 <rwhitby-web> only affects the two Makefiles then, and doesn't invalidate any of the beta candidate testing
May 10 23:23:21 <jbowler> ;-) well, my hack works.... but it will download everything from www.oesources.org because SOURCEFORGE_MIRROR fails to provide...
May 10 23:24:34 <jbowler> The problem with flex is that the nslu.sf.net download is not in a 'lex' subdirectory, which is what the .bb file requires
May 10 23:25:03 <jbowler> The problem with everything else seems to be that there is no additional searching going on.
May 10 23:25:22 <rwhitby-web> Just get the top-level Makefile to download it if it is not already in the downloads dir, and make that a dependency of the firmware target.
May 10 23:27:42 <jbowler> Yes, that sounds much better (though my fix was in the makefile too - I just changed local.conf).  How to get it?  wget?
May 10 23:28:30 <rwhitby-web> yep
May 10 23:30:11 <ka6sox-away> thats what I used and it works fine and builds correctly.
May 10 23:30:18 <rwhitby-web> I love how ssh can bounce the authentication agent through three hops (here to home to nudi to berlios)
May 10 23:30:34 <ka6sox-away> jbowler, thanks for staying up and fixing this up.
May 10 23:30:34 <jbowler> The MD5 in the nslu.fs.net/downloads directory is 0 sized
May 10 23:30:46 <ka6sox-away> ignore it(the build does)
May 10 23:31:28 <rwhitby-web> I've removed it
May 10 23:31:36 <jbowler> How about putting the correct one there?  I don't know if not downloading it will be ok, and downloading something bogus sounds bad.
May 10 23:31:55 <rwhitby-web> not downloading it is fine for everything else we have there
May 10 23:31:58 <ka6sox-away> I'm not sure where to find a correct one.
May 10 23:32:18 <ka6sox-away> since the mirrors for prdownloads.sourceforge.net are bad
May 10 23:32:20 <rwhitby-web> the .md5 is generated locally, not downloaded
May 10 23:32:27 <ka6sox-away> ah.
May 10 23:37:50 <jbowler> Curious, I added the makefile dep and it still tries to contact aleron, but now, because the makefile dl's it, it works ok.  I'm worried because my failure case does not produce the .bz2 file (it produces some bogus HTML file), but it sure seems to work ;-)
May 10 23:39:09 <ka6sox-away> are we doing the same thing for Unslung?
May 10 23:40:06 <jbowler> That's what I tested it on, just trying to work out how to put svn openslug back to a 'distclean' state
May 10 23:43:22 <ka6sox-away> I don't see a target for clobbering the downloads folder.
May 10 23:43:34 <rwhitby-web> correct
May 10 23:44:22 <rwhitby-web> I will add a distclean
May 10 23:44:37 <jbowler> I've got one already.
May 10 23:45:11 <jbowler> Should I commit this stuff (wget, added firmware target, distclean removes everything including the flex arch)
May 10 23:46:08 <jbowler> If it gets past flex-native it is safe, because it just pre-populates the downloads folder.
May 10 23:46:09 <rwhitby-web> yeah, ok.  Then svn cp it over to the release area for both too
May 10 23:46:36 <ka6sox-away> thanks guys.
May 10 23:46:40 <jbowler> Um... I made the change in the 'release' area - this is bad?
May 10 23:46:46 <rwhitby-web> distclean removes downloads completely too?
May 10 23:46:58 <rwhitby-web> make it in both the release and the trunk using svn cp
May 10 23:47:13 <jbowler> No, distclean just removes the specific flex archive, should it rm -rf downloads?
May 10 23:47:19 <rwhitby-web> yes
May 10 23:47:40 <rwhitby-web> so clobber just removes tmp, and distclean removes everything that was not put there by svn
May 10 23:48:34 <jbowler> ...not put there by make or (if I got it right) bitbake...
May 10 23:49:52 <jbowler> svn cp: is it valid to make a change in release/*/Makefile then svn cp release/*/Makefile trunk/*/Makefile, or must I do it the other way round?
May 10 23:50:31 <rwhitby-web> either is fine
May 10 23:50:32 <jbowler> silly question, I can just swap the files then do it the way I know to be correct ;-)
May 10 23:51:36 <rwhitby-web> distclean removes everything which *was* put there by make or bitbake
May 10 23:51:54 <rwhitby-web> which is everything which was *not* put there by svn
May 10 23:52:48 <jbowler> Right.  So the externals, like 'bitbake' and 'nslu2-linux' stay, tmp (clobber), downloads, setup-env and conf/local.conf go
May 10 23:53:10 <rwhitby-web> correct
May 10 23:53:36 <rwhitby-web> so if you svn co, then make, then make distclean, the dir should be identical to just an svn co
May 10 23:55:16 <jbowler> I believe it will be ('cause bitbake only creates in tmp and download and I verified what the Makefile makes.)
May 10 23:55:34 <jbowler> svn cp releases/*/Makefile trunk/*/Makefile
May 10 23:55:34 <jbowler> svn: File 'trunk/openslug/Makefile' already exists
May 10 23:56:14 <jbowler> I can do a simple cp and commit both, but this sounds slightly weird.
May 10 23:56:51 <rwhitby-web> commit to trunk, then use a remote cp to copy it (use the svn ssh://..... URLS for both pathnames)
May 10 23:57:57 <rwhitby-web> svn cp -m "Updated Makefile in release" svn ssh://svn.berlios.de/svnroot/repos/openslug/trunk/openslug/Makefile svn ssh://svn.berlios.de/svnroot/repos/openslug/releases/OpenSlug-1.10-beta/Makefile
May 10 23:58:09 <rwhitby-web> (put in the missing plus signs that this web interface eats)
May 10 23:59:04 <rwhitby-web> see svn help cp and the "complete server side copy"
**** ENDING LOGGING AT Tue May 10 23:59:56 2005