**** BEGIN LOGGING AT Sun Jun 12 23:59:57 2005
Jun 13 00:15:04 <rwhitby> still only 5 entries in the 5.x alpha tester database: http://groups.yahoo.com/group/nslu2-linux/database?method=reportRows&tbl=10
Jun 13 00:15:23 <rwhitby> has no-one else here tried it?  This is supposed to be the developer channel .....
Jun 13 00:21:39 <rwhitby> kared: you loaded up 5.x didn't you?
Jun 13 00:21:52 <rwhitby> jacques: you loaded 5.x too
Jun 13 00:27:40 <jf-work> Gmorning
Jun 13 00:27:50 <rwhitby> g'day jf-work
Jun 13 00:28:00 <rwhitby> you got a slug to try out 5.5-beta?
Jun 13 00:28:25 <jf-work> rwhitby: I've got a poduction one, yep
Jun 13 00:28:36 <jf-work> s/poduction/production
Jun 13 00:35:26 <kared> rwhitby: im running openslug
Jun 13 00:40:28 <rwhitby> kared: ah - I must be thinking of someone else ...
Jun 13 00:40:41 <rwhitby> ah - that's right, it was "ilg"
Jun 13 01:07:36 <jf-work> anyone using cups optware ? Everything builds one wiley except pdftops. I was wondering if this is important or not
Jun 13 01:12:59 <jf-work> s/one/on
Jun 13 01:52:14 <jacques> problem with unslung-5.5 build:
Jun 13 01:52:40 <jacques> it downloads flex at the beginning, but it assumes there's a downloads dir (and there isn't)
Jun 13 01:52:57 <jacques> luckily, flex downloads properly again from the real source
Jun 13 01:53:28 <jacques> cd downloads; wget -nc http://nslu.sourceforge.net/downloads/flex-2.5.31.tar.bz2
Jun 13 01:53:28 <jacques> /bin/sh: line 1: cd: downloads: No such file or directory
Jun 13 01:53:48 <jacques> so it ends up downloading flex into the root build dir
Jun 13 02:07:11 <rwhitby> jacques: so you're saying we can remove that target from the makefile?
Jun 13 02:07:18 <rwhitby> OE will create the dir properly I think
Jun 13 02:11:41 <jacques> oe creates the dir
Jun 13 02:11:48 <jacques> but it's too late for that flex binary
Jun 13 02:11:53 <jacques> source I mean
Jun 13 02:12:16 <jacques> so we should probably just remove it from the makefile as you said
Jun 13 02:35:16 <rwhitby> yeah, doing that now.  unfortunately that means having to check out the whole svn repo (trunk and releases) and that takes a while
Jun 13 02:39:52 <jacques> my Unslung-5.5-beta build completed
Jun 13 02:40:04 <jacques> yay, everything downloads properly
Jun 13 02:40:13 <rwhitby> that must be a first ....
Jun 13 02:40:31 <jacques> I think it is a first since we started using svn
Jun 13 02:52:41 <dyoung-zzzz> It doesnt build for me.
Jun 13 02:52:47 <dyoung-zzzz> I get:
Jun 13 02:52:52 <dyoung-zzzz> ERROR: myProgressCallback() takes exactly 3 arguments (5 given) while parsing (some .bb file)
Jun 13 02:53:17 <dyoung-zzzz> (as it parses all the .bb files)
Jun 13 02:54:59 <jacques> you still have a PYTHONPATH set or something ?
Jun 13 03:08:07 <CIA-12> 03rwhitby * r17 10/trunk/unslung/Makefile: Removed the work-around for flex.
Jun 13 03:08:39 <CIA-12> 03rwhitby * r18 10/releases/Unslung-5.5-beta/Makefile: Removed the work-around for flex.
Jun 13 03:08:59 <rwhitby> jacques: fixed
Jun 13 03:09:17 <rwhitby> won't affect the binary.
Jun 13 03:10:51 <jacques> I agree
Jun 13 03:11:06 <rwhitby> my build just finished
Jun 13 03:11:23 <rwhitby> (but I cheated and symlinked the downloads dir)
Jun 13 03:11:50 <dyoung-zzzz> jacques, the PYHTONPATH was the fix for that particular problem.  Now I get some ugly tracebacks from python.
Jun 13 03:12:11 <dyoung-zzzz> do_fetch started, the immediate tracebacks
Jun 13 03:12:57 <rwhitby> dyoung-zzzz: do you have any trace of your installed bitbake? (cause we updated bitbake again to version 250).
Jun 13 03:13:26 <dyoung-zzzz> its referenceing the right place .../unslung/bitbake/lib/bb/foo.py.
Jun 13 03:13:35 <dyoung-zzzz> (I uninstalled the installed one)
Jun 13 03:15:41 <dyoung-zzzz> NOTE: package quilt-native-0.39-r0: task do_fetch: started
Jun 13 03:15:41 <dyoung-zzzz> Traceback (most recent call last):
Jun 13 03:15:41 <dyoung-zzzz>   File "/home/dereky/unslung/bitbake//bin/bitbake", line 720, in ?
Jun 13 03:15:48 <dyoung-zzzz> if that helps any....
Jun 13 03:15:58 <jacques> dyoung-zzzz, no BBPATH set either right?
Jun 13 03:16:40 <dyoung-zzzz> Nope, I logged out and logged back in for good measure.
Jun 13 03:17:04 <dyoung-zzzz> just svn co svn://foo; cd unslung; make
Jun 13 03:19:39 <rwhitby> Beep: Firmware Version:  V2.3R29-uNSLUng-5.5-beta   
Jun 13 03:19:49 <jacques> cool
Jun 13 03:20:06 <dyoung-zzzz> I check out the truni; I'll try the release now just in cas.
Jun 13 03:21:57 <rwhitby> dyoung-zzzz: if we get this binary release out in the next day or so, will we be back on the release periodicity track after the slow 4.x schedule?
Jun 13 03:22:34 <dyoung-zzzz> I dunno.  I'm not so sure theres that much more to do.
Jun 13 03:22:55 <dyoung-zzzz> just kernel modules and stuff; but that can just be in the feed.
Jun 13 03:23:30 <rwhitby> so 5 releases divided by how many months?
Jun 13 03:23:41 <dyoung-zzzz> 10.5 or so.
Jun 13 03:23:56 <rwhitby> so back on track for every two months
Jun 13 03:24:24 <dyoung-zzzz> I mean if someone has some Super Suggestion for kernel stuffs; I'm sure we can keep the schedule.
Jun 13 03:24:29 <dyoung-zzzz> well, the release works.
Jun 13 03:24:32 <dyoung-zzzz> the trunk doesnt work for me.
Jun 13 03:24:58 <dyoung-zzzz> I blew away temp (foro the trunk builds) just inc ase there was something busted there
Jun 13 03:25:11 <dyoung-zzzz> I suppose it doesnt matter as long as the release one works.
Jun 13 03:25:30 <dyoung-zzzz> and I'm letting it download everything too; so this is gonna take a while.
Jun 13 03:26:06 <rwhitby> dyoung-zzzz: found it
Jun 13 03:26:28 <dyoung-zzzz> was it user error?
Jun 13 03:26:53 <rwhitby> svn:externals on trunk is wrong.  release is ok
Jun 13 03:27:09 <dyoung-zzzz>  ok
Jun 13 03:27:21 <rwhitby> fixing now
Jun 13 03:27:22 <dyoung-zzzz> I'm not going to worry about it then.
Jun 13 03:27:59 <rwhitby> wait a minute, and then try trunk again.  no sense in all of us testing the release branch and no-one testing the trunk
Jun 13 03:28:31 <dyoung-zzzz> ok
Jun 13 03:28:47 <dyoung-zzzz> I'll interrupt the release when yourue ready.
Jun 13 03:28:49 <CIA-12> 03rwhitby * r19 10/trunk/unslung/: Updated svn:externals to fix bitbake
Jun 13 03:28:56 <rwhitby> go for it now
Jun 13 03:29:49 <jacques> Updated external to revision 250.
Jun 13 03:29:58 <jacques> in trunk
Jun 13 03:30:19 <rwhitby> that'll fix the bitbake errors
Jun 13 03:31:12 <jacques> is it useful for me to flash unslung 5.5 and see if it boots? because I need that slug for several hours very soon
Jun 13 03:33:43 <dyoung-zzzz> flash/boot it so you can rightly add your name to the database. :-)
Jun 13 03:34:10 <dyoung-zzzz> Ok, the trunk build is going.
Jun 13 03:34:20 <jacques> `k
Jun 13 03:35:01 <dyoung-zzzz> I expect this build to last longer than me.
Jun 13 03:35:29 <dyoung-zzzz> anything else that needs looking at today?
Jun 13 03:36:16 <rwhitby> just the README
Jun 13 03:40:28 <dyoung-zzzz> so have we elected to leave the stock linksys password alone?
Jun 13 03:41:03 <rwhitby> yep
Jun 13 03:41:09 <dyoung-zzzz> ie: notsling the existing rootpassword on the candidate unslingign disc.
Jun 13 03:41:11 <dyoung-zzzz> okay
Jun 13 03:41:48 <jacques> V2.3R29-uNSLUng-5.5-beta
Jun 13 03:41:50 <jacques> beep
Jun 13 03:42:16 <jacques> hmm, if unslung can restore the timezone, why can't openslug?  :-\
Jun 13 03:43:08 <dyoung-zzzz> I havnt actually lfollowed the README to the letter in installation 9since I'm not done building it yet), but it looks tood to me.
Jun 13 03:43:58 <jacques> I added myself to the yahoo db
Jun 13 03:44:52 <dyoung-zzzz> jacques, thanks; we're one close to beta.
Jun 13 03:44:57 <dyoung-zzzz> "true beta"
Jun 13 03:45:15 <jacques> yep
Jun 13 03:48:09 <jacques> now that little sluggie has to do something else for a few hours
Jun 13 03:49:23 <dyoung-zzzz> Test!
Jun 13 03:49:25 <dyoung-zzzz> Test!
Jun 13 03:49:31 <dyoung-zzzz> Qualify!
Jun 13 03:49:40 <dyoung-zzzz> Go Little Sluggie, Go!
Jun 13 03:50:01 <dyoung-zzzz> I wonder how many days it would take a slug to self-build a kernel.
Jun 13 03:50:15 <jacques> two hours
Jun 13 03:50:30 <dyoung-zzzz> thats it?
Jun 13 03:50:40 <dyoung-zzzz> I figured the bitbake parsing alone would take a couple hours.
Jun 13 03:50:42 <jacques> is my guess based on building kernels on other arm boxes
Jun 13 03:50:53 <jacques> you didn't mention bitbake
Jun 13 03:50:57 <jacques> :-\
Jun 13 03:51:09 <dyoung-zzzz> Okay, let me qualify that.
Jun 13 03:51:14 <jacques> I don;t think a 32MB slug can realistically run bitbake
Jun 13 03:51:24 <dyoung-zzzz> "I wonder how long it would take a slug to bitbake itself"
Jun 13 03:51:30 <jacques> oh, the two hours is for 2.4 - longer for 2.6
Jun 13 03:51:59 <dyoung-zzzz> (I'm not planning to actually do it)
Jun 13 03:53:02 <dyoung-zzzz> it takes long enough to bitbake unslung-image on a P4 3Ghz.
Jun 13 03:53:24 <jacques> at least ehe memory usage is reasonable now
Jun 13 03:53:48 <dyoung-zzzz> Yeah.  A FatSlug should be able to handle it.
Jun 13 03:55:41 <jacques> yeah
Jun 13 03:56:00 <jacques> hmm, wanna hear something slightly alarming? (at least to me)
Jun 13 03:56:41 <dyoung-zzzz> sure.
Jun 13 04:05:31 <rwhitby> ...
Jun 13 04:06:17 <jacques> rwhitby, sorry I decided it was better brought up in another channel...
Jun 13 04:06:36 <jacques> which you are also on
Jun 13 04:06:53 <rwhitby> gotcha
Jun 13 04:11:30 <rwhitby> ok, doesn't affect the beta ...
Jun 13 04:12:59 <jacques> nope
Jun 13 04:32:32 <rwhitby> dyoung-zzzz: do we still need portmap-unslung in OE ?
Jun 13 04:34:37 <rwhitby> glc: is portmap-unslung needed for nfsroot in linuxrc?  I can't see it used in there.
Jun 13 04:37:31 <rwhitby> ilg: have you flashed 5.x?
Jun 13 04:38:36 <ilg> 5.3
Jun 13 04:38:58 <rwhitby> we have a 5.5-beta available as an svn source release if you want to try that.
Jun 13 04:39:24 <rwhitby> Can you also please add your entry to http://groups.yahoo.com/group/nslu2-linux/database?method=reportRows&tbl=10
Jun 13 04:40:39 <dyoung-zzzz> we dont need it *if* folks dont wish to run the kernel nfs with locking.
Jun 13 04:41:01 <dyoung-zzzz> if locking is desired, we still need it.
Jun 13 04:41:29 <rwhitby> dyoung-zzzz: but the stock linuxrc doesn't use it, right?
Jun 13 04:42:00 <dyoung-zzzz> It shouldnt need it, as the recommended way to nfs rootfs mount is with nolock option.
Jun 13 04:42:33 <rwhitby> recommended for linux in general, or just for nslu2-linux?
Jun 13 04:42:55 <dyoung-zzzz> nslu2-linux.
Jun 13 04:44:59 <ilg> rwhitby: I followed your link and joined the Yahoo group nslu2-linux. is this what you suggested?
Jun 13 04:45:27 <rwhitby> and added your entry to the database?
Jun 13 04:46:14 <rwhitby> we usually don't release a binary until we get 20 alpha testers, but we'll probably make an exception this time and do it on 10.
Jun 13 04:47:35 <ilg> where do I get the unslung#?
Jun 13 04:47:52 <rwhitby> just use the next number
Jun 13 04:49:40 <ilg> done. I'll get 5.5 and give it a try later
Jun 13 04:49:58 <dyoung-zzzz> upslugging 5.5
Jun 13 04:50:28 <rwhitby> ilg: thanks!
Jun 13 04:51:18 <rwhitby> ilg: there are significant differences between 5.3 and 5.5, but most are just removal of legacy stuff, so that won't affect you.
Jun 13 04:51:58 <rwhitby> we're missing ka6sox's entry from the table too
Jun 13 04:52:04 <ilg> linksys sources question: I would like to patch the web interface to accept workgroup names containing '.', and I located the file Management/setup.cgi. do we have access to the sources of it? any other suggestions on how to solve this?
Jun 13 04:52:17 <rwhitby> we don't have the sources to that
Jun 13 04:54:43 <dyoung-zzzz> beep! ; V2.3R29-uNSLUng-5.5-beta                                                        
Jun 13 04:54:52 <ilg> :-(  I did not check openslug yet, any chance to configure my workgroup name with it?
Jun 13 04:55:32 <rwhitby> ilg: if you do a strings on setup.cgi, you will see a number of lists of punctuation chars.  Maybe doing a binary sed on setup.cgi to replace the '.' with one of the other chars will fix it?
Jun 13 04:55:54 <rwhitby> or just use a diversion script to edit the samba.conf file after Linksys is finished with it.
Jun 13 04:56:12 <dyoung-zzzz> ilg, have you tried editing /etc/CGI_ds.conf and rebooting ?
Jun 13 04:56:13 <rwhitby> or use the samba version 3 Unslung package
Jun 13 04:56:39 <rwhitby> yeah, lots of options
Jun 13 04:56:51 <ilg> yes, a lot to choose from :-)
Jun 13 04:57:06 <ilg> I'll investigate some and let you know
Jun 13 04:57:40 <rwhitby> ilg: when you find the best one, please add a HowTo/UseADotInTheWorkgroupName wiki page
Jun 13 04:58:13 <ilg> ok
Jun 13 07:21:15 <rwhitby> Unslung 5.5-beta source release has been announced.
Jun 13 08:39:00 <CIA-12> 03bzhou * 10unslung/make/py-bittorrent.mk: upstream upgrade from 4.1.1 to 4.1.2
Jun 13 08:42:45 <ka6sox-zzzz> Unslung-5.5-beta binaries are available at www.unslung.org
Jun 13 08:55:33 <ka6sox> rebooted Boo into linux and upslugged to devslug
Jun 13 08:55:39 <ka6sox> V2.3R29-uNSLUng-5.5-beta
Jun 13 08:55:56 <jacques> ka6sox you in the database yet? :-)
Jun 13 08:56:08 <ka6sox> just added #8
Jun 13 08:56:29 <jacques> wow, two more and we're golden
Jun 13 08:57:52 <ka6sox> anyone who downloads it and sucessfully installs it please add your info to the yahoo database.
Jun 13 09:00:12 <jbowler> rwhitby-asleep: reading the unslung linuxrc it looks like the main problem is that the ethernet id (MAC) isn't set.  There's a piece of devio stuff at line 86 of the latest sysconfsetup (search for 02X) which will work on unslung - it reads the MAC from RedBoot.
Jun 13 09:01:17 <jbowler> rwhitby-asleep: on openslug /boot/network sets the MAC (in fact this is unnecessary because ifup does it too).  The insmod (etc) isn't necessary on openslug because the modprobe from the kernel works.
Jun 13 09:02:52 <jacques> can we make openslug restore the TZ on firstboot like unslung does?
Jun 13 09:03:41 <jbowler> jacques: set it in /etc/default/rcS
Jun 13 09:04:22 <jacques> I've been installing a tzfile - scping /etc/localtime from one of my desktops
Jun 13 09:04:38 <[g2]> we need to start a list of all the sensible things to keep track of
Jun 13 09:04:39 <jbowler> Um, no, there's somewhere else.  I think /etc/timezone?  It should be necessary to install a complete tz setup.
Jun 13 09:04:48 <jbowler> s/should/shouldn't/
Jun 13 09:05:10 <jacques> /etc/localtime thing works
Jun 13 09:06:00 <jbowler> Right, but I thought I got it to work just by setting a simple ASCII string somewhere, tzfile requires a program to generate doesn't it?
Jun 13 09:06:25 <jacques> jbowler, yes, I admit my way is not the best - I am having to use the binary tzfile from my desktops
Jun 13 09:06:48 <jbowler> Good thing they got the byte sex issues right there ;-)
Jun 13 09:06:50 <jacques> but I wonder why gentoo does not use the simple ascii method
Jun 13 09:07:30 <[g2]> could you guys explain what I'm missing be just exporting TZ=EST5EDT4 ?
Jun 13 09:07:40 <jbowler> Because of the DST algorihtms I believe.
Jun 13 09:07:45 <[g2]> s/be/by/
Jun 13 09:07:54 <jacques> I suspected it had something to do with DST
Jun 13 09:08:40 <[g2]> I'm still missing the boat
Jun 13 09:09:48 <[g2]>  echo $TZ && date
Jun 13 09:09:48 <[g2]> EST5EDT4
Jun 13 09:09:48 <[g2]> Mon Jun 13 12:09:36 EDT 2005
Jun 13 09:10:07 <[g2]> that looks right to me
Jun 13 09:10:52 <CIA-12> 03bzhou * 10unslung/ (Makefile make/hexcurse.mk): hexcurse ready for cross testing
Jun 13 09:13:25 <jacques> [g2], I dunno what the difference is really
Jun 13 09:14:55 <[g2]> ok
Jun 13 09:15:21 <[g2]> I don't either that's why I asked :)
Jun 13 09:16:27 <jacques> perl build finished
Jun 13 09:16:36 <jacques> running testsuite now
Jun 13 09:16:50 <[g2]> that's the native build ?
Jun 13 09:17:07 <jacques> native glibc-2.3.5 build :-)
Jun 13 09:17:29 <[g2]> cool
Jun 13 09:17:38 <[g2]> are you running 2.3.5 with openslug ?
Jun 13 09:17:52 <jacques> yep
Jun 13 09:18:19 <[g2]> so where is that magic chant that sets the glibc provider ?
Jun 13 09:18:36 <jacques> just a sec - finding it
Jun 13 09:19:03 <jacques> PREFERRED_VERSION_glibc = "2.3.5+cvs20050612"
Jun 13 09:19:03 <jacques> CVSDATE_glibc = "20050612"
Jun 13 09:19:29 <jacques> hmm, tho you may want to change that cvsdate
Jun 13 09:19:33 <[g2]> that's pretty crispy fresh eh ? :)
Jun 13 09:20:19 <jacques> pb just committed the glibc-fp-byteorder.patch to glibc cvs this morning - I had to do it manually
Jun 13 09:20:57 <[g2]> you just added those to you local.conf ?
Jun 13 09:21:44 <jacques> yeah
Jun 13 09:22:02 <[g2]> do you see any reason not to go to the latest ?
Jun 13 09:22:11 <jacques> latest ?
Jun 13 09:22:34 <[g2]> well 2.3.5 2005 06 12 would seem like the latest to me :)
Jun 13 09:22:54 <jacques> I built it yesterday
Jun 13 09:22:57 <[g2]> that's yesterday's cvs snapshot
Jun 13 09:23:12 <jacques> and since I was building more packages today, I had to add the CVSDATE_glibc
Jun 13 09:23:54 <[g2]> my question is which glibc do you think openslug should be baselined on
Jun 13 09:24:21 <jacques> oooh
Jun 13 09:24:40 <jacques> I consider this still in testing
Jun 13 09:24:46 <jacques> but it's looking good so far
Jun 13 09:25:01 <[g2]> OK, thx
Jun 13 09:25:19 <[g2]> at some point we'll probably want to change
Jun 13 09:25:29 <jacques> when we update openslug glibc, any version after the patch pb applied should be fine as long as nothing else breaks later
Jun 13 09:25:54 <jacques> I dunno what the plans are for 2.3.5 official release
Jun 13 09:26:15 <[g2]> right, as long as OE doesn't switch to 2.3.5 underneath us
Jun 13 09:26:45 <jacques> I don't see that happening soon - nobody's even really using 2.3.5 on armel yet from what ppl said today in #eo
Jun 13 09:26:48 <jacques> #oe
Jun 13 09:26:57 <kergoth> ?  just set your PREFERRED_VERSION if you dont want to use whatever the default is
Jun 13 09:27:02 <kergoth> heh
Jun 13 09:27:21 <[g2]> morning sunshine !
Jun 13 09:30:20 <[g2]> kergoth, did you here the APEX news ?
Jun 13 09:30:30 <kergoth> nope
Jun 13 09:30:51 <[g2]> beewoolie got the kernel to load from the jffs2 partition last night
Jun 13 09:31:26 <[g2]> and I've been running APEX with a 64MB modded slug for a couple weeks now
Jun 13 09:39:19 <kergoth> nice
Jun 13 10:05:18 <marceln>  /join #nslu2-general
Jun 13 10:43:02 <jacques> All tests successful. - glibc 2.3.5 looks good
Jun 13 10:43:45 <ka6sox> for openslug?
Jun 13 10:45:24 <jacques> perl/openslug/glibc-2.3.5
Jun 13 10:45:46 <[g2]> jacques, great!
Jun 13 10:49:56 <VoodooZ_Log> great stuff
Jun 13 12:21:07 <jbowler> [g2]: I assigned you slugbug #145, because it's the one for the shutdown -r now problem.
Jun 13 12:22:42 <[g2]> jbowler, ok
Jun 13 12:52:11 <jbowler> [g2]: have you seen slugbug #109?  It might be the RTC issue causing the hang, but it's worrying that it seems associated with high disk activity/irq 26
Jun 13 12:54:20 <[g2]> jbowler, I don't recall #109
Jun 13 12:56:22 <jbowler> It's just a duplicate of the irq26 report I think.  The other thing in there was on a different system and sounds like a description of what would happen on out of memory.
Jun 13 13:12:23 <jbowler> There is no report of the irq26 problem that I can find, apart from #109.
Jun 13 13:16:28 <CIA-12> 03jeanfabrice * 10unslung/Makefile: Promote gdchart, hexcurse for wiley
Jun 13 13:30:08 <jeanfabrice> bob_tm-away: ping
Jun 13 13:31:29 <bob_tm-away> jeanfabrice: Pong... Gotta reboot (XP getting problematic - back in 3-4 mins)
Jun 13 13:31:38 <jeanfabrice> ok, let's reboot
Jun 13 13:36:32 <jeanfabrice> re bob_tm
Jun 13 14:47:04 <NAiL> ka6sox: Easiest way to mirror? rsync? :)
Jun 13 14:48:25 <ka6sox-office> yes.
Jun 13 14:48:51 <ka6sox-office> NAiL, finally back in office and packing up your cable today.
Jun 13 14:49:06 <ka6sox-office> so you should see it in about a week.
Jun 13 14:49:22 <NAiL> ka6sox-office: Cool. I'll try to get my damn CC activated so I can pay through paypal. 
Jun 13 14:49:58 <NAiL> I've been so busy lately, so I haven't exactly had time to tinker with stuff anyways
Jun 13 14:50:10 <ka6sox-office> me niether.
Jun 13 14:50:24 <ka6sox-office> just had time this weekend to work on releases.
Jun 13 15:22:14 <jacques> and I the only person getting those damn dropbear disconnects?
Jun 13 15:22:18 <jacques> in openslug
Jun 13 15:23:16 <[g2]> jacques, I haven't gotten any but then I only keep the dropbear connection up for 6-8 hours at a shot
Jun 13 15:23:35 <[g2]> are you running IPV6 by chance ?
Jun 13 15:24:38 <jacques> no ipv6
Jun 13 15:24:57 <jacques> it does seem to happen after more than 8 hours
Jun 13 15:25:15 <jacques> ah, well, openssh works much better in so many ways
Jun 13 15:25:26 <jacques> but it's a shame because that didn't used to happen with dropbear
Jun 13 15:25:56 <jacques> and the very long timescale makes it hard to debug
Jun 13 15:26:09 <jacques> can't reproduce in less than 10-12-more hours
Jun 13 15:26:55 <NAiL> ka6sox-office: Everything is set up. As soon as I get the key and name, it's up after two secs ;)
Jun 13 15:27:24 <ka6sox-office> NAiL, thanks. emailing you the info now.
Jun 13 15:27:50 <[g2]> jacques, is there a timeout in dropbear ?
Jun 13 15:27:56 <NAiL> ka6sox-office: oh, ok :)
Jun 13 15:28:12 <[g2]> I mean an activity timeout
Jun 13 15:28:32 <NAiL> [g2]: I've had dropbear sessions running for ages without any problems with no activity. But I know dropbear has been updated since then...
Jun 13 15:29:15 <ka6sox-office> but don't stay up for it as I may get distracted and not get it out now.
Jun 13 15:29:37 <jacques> [g2], if there is it's a new "feature"
Jun 13 15:30:50 <[g2]> well they did just change the login with no passwd to make it more secure right ?
Jun 13 15:32:34 <NAiL> ka6sox-office: Nah, probably going to watch a movie or something anywas
Jun 13 15:32:37 <NAiL> anyways
Jun 13 15:32:42 <ka6sox-office> j
Jun 13 15:32:45 <ka6sox-office> er k
Jun 13 16:33:19 <rwhitby-web> when I read the binary announcement email in Gmail, the three sponsored links are: 1) BSA "Copying software illegally is stealing. Report software piracy!", 2) Microsoft "Get The Facts: Windows vs. Linux. Read The Independent Analysis Now." and 3) ABC LanLicenser "Keep your software under control, stay legal, reduce your costs."
Jun 13 16:35:41 <Tiersten> heh
Jun 13 19:29:48 <NoRemorse> hi, can anyone tell me if the "disk backup" function of the standard firmware creates an identical copy of disk 1? ie can the backup disk then be swapped to port 1 and all config and shares etc will work the same?
Jun 13 19:38:05 <NoRemorse> surely someone here has used the backup to disk function before?
Jun 13 19:44:25 <NAiL> hmm
Jun 13 19:45:04 <NAiL> Haven't tried. Actually I can't say I've used the original firmware. But I'd guess it's not a very bright idea though...
Jun 13 19:46:58 <NoRemorse> they seem to have NO support for a "restore" so they only thing I can think of is the backup disk becomes a direct plugin replacement for slot 1
Jun 13 22:48:50 * VirtualCtor is overwhelmed at the feverish pace of everyone working quietly to themselves
Jun 13 22:49:15 <ka6sox-away> :)
Jun 13 22:50:50 <rwhitby-web> VirtualCtor: we're all just waiting for the 5.x bug reports ....
Jun 13 22:51:28 <VirtualCtor> Oh, then I take it all back. :)
Jun 13 22:54:17 <VirtualCtor> Actually, I stopped in to discuss an idea for change in the startup process.
Jun 13 22:54:48 <VirtualCtor> Maybe somebody can tell me it's a stoopid idea.
Jun 13 23:02:09 <ka6sox-away> go ahead
Jun 13 23:02:47 <VirtualCtor> The init scripts in /opt/etc/init.d are all prefixed with the S*.  Most of the time, these S* or K* scripts are in the rc*.d directories and are links to the actual scripts in the init.d directory.  Is there some reason that this is different in unslung?
Jun 13 23:03:45 <ka6sox-away> there is only one runlevel on a slug.
Jun 13 23:04:07 <ka6sox-away> why have all that extra junk hanging around (with links etc)
Jun 13 23:04:25 <mr_claus> morning
Jun 13 23:04:33 <ka6sox-away> morning
Jun 13 23:05:19 <mr_claus> i want to build a new package, the install step which it called with "make install" try to change the UID which is not possible if running with another user as root
Jun 13 23:05:28 <VirtualCtor> Because, it makes it easier to turn services on and off if it's just a link.
Jun 13 23:05:30 <mr_claus> how that is solved with other packages
Jun 13 23:05:57 <ka6sox-away> mr_claus: look at Screen
Jun 13 23:05:57 <mr_claus> should i patch the makefile or copy the files manually to their destination?
Jun 13 23:06:16 <VirtualCtor> That way, you don't have to modify the script, move the script around, etc.
Jun 13 23:06:23 <ka6sox-away> VirtualCtor, put that suggestion into the Unslung wishlist on the wiki
Jun 13 23:06:50 <ka6sox-away> ~nslu2-wiki
Jun 13 23:06:51 <jbot> extra, extra, read all about it, nslu2-wiki is the short-hand way of saying "Have a look at the NSLU2-Linux Wiki at http://www.nslu2-linux.org/".
Jun 13 23:06:53 <mr_claus> ka6sox-away: hmm?
Jun 13 23:07:19 <ka6sox-away> mr_claus, look at the way that problem is solved in the Screen package
Jun 13 23:07:19 <VirtualCtor> Okay, let me look it over for a bit...
Jun 13 23:07:39 <mr_claus> ka6sox-away: ok, thx
Jun 13 23:10:43 <mr_claus> ka6sox-away: ok, files will be copied by hand, thats only a solution for packages with less files, but the important question for me is will the uid changed while the files are being packed into the ipkg-file?
Jun 13 23:12:58 <ka6sox-away> if you set them ipkg should preserve them.
Jun 13 23:13:17 <ka6sox-away> (otherwise a LOT of things would break)
Jun 13 23:15:30 <ka6sox-away> dpkg in Debian and RPM in Redhat preserve permissions (and iirc ipkg is based on dpkg)
Jun 13 23:16:15 <mr_claus> ka6sox-away: oh, i see that permissions are being set while installing the package with the postinst step
Jun 13 23:19:53 <ka6sox-away> yep
Jun 13 23:20:26 <ka6sox-away> since the package manager runs as a priveleged process.
Jun 13 23:21:05 <mr_claus> but how the dhcp package will do it i.e.?
Jun 13 23:21:15 <mr_claus> i dont see a postinst step
Jun 13 23:21:33 <mr_claus> the binary in the build path is not owned by root
Jun 13 23:21:41 <mr_claus> but the installed binary is owned by root
Jun 13 23:22:03 <ka6sox-away> its in /usr/bin?
Jun 13 23:23:21 <ka6sox-away> mr_claus, is this for NSLU2 or WL500g?
Jun 13 23:24:56 <mr_claus> its for NSLU2
Jun 13 23:25:06 <mr_claus> and dhcp is in /opt/sbin
Jun 13 23:25:33 <ka6sox-away> can you tell me what package this is? (or did I miss it earlier?)
Jun 13 23:27:43 <mr_claus> make/dhcp.mk
Jun 13 23:28:50 <ka6sox-away> okay so you are working on the Daemon.
Jun 13 23:28:52 <ka6sox-away> cool.
Jun 13 23:29:23 <mr_claus> oh, thats a sample, i know it because i have done a change a few days ago
Jun 13 23:29:45 <ka6sox-away> your package is?
Jun 13 23:29:46 <mr_claus> now i want to include the libnss-ldap package, it's ready to build but only as user root yet
Jun 13 23:29:58 <ka6sox-away> okay...understood.
Jun 13 23:30:24 <mr_claus> the ldap replication from the provider-server is working, now i have to integrate libnss-ldap to get ldap running with samba and the filesystem
Jun 13 23:31:17 <ka6sox-away> okay understood.
Jun 13 23:31:22 <mr_claus> next step would be acl's but thats not very simple because a lot of packages need to be change and i dont know if it makes sense on unlung
Jun 13 23:31:53 <ka6sox-away> send an email asking the question on the ML and one of the package managers will respond (when they have time)
Jun 13 23:32:39 <mr_claus> oh, thats a funny thing, i tried to get access rights to the mailing list a few days ago but i never got the mail that i'm at the list now
Jun 13 23:32:59 <ka6sox-away> hmmm...what is your ID?
Jun 13 23:33:06 <ka6sox-away> (yahoo ID)
Jun 13 23:33:52 <mr_claus> i have none ID, i tried to register but it never worked
Jun 13 23:34:11 <ka6sox-away> ah..try re-registering.
Jun 13 23:36:53 <mr_claus> yes, i try
Jun 13 23:39:44 <VirtualCtor> ka6sox-away: okay, I added the suggestion.  Let me know if I need to make it clearer.
Jun 13 23:39:55 <ka6sox-away> k
Jun 13 23:40:50 <mr_claus> hmm, no mail
Jun 13 23:41:08 <ka6sox-away> what is your email addy?
Jun 13 23:41:19 <ka6sox-away> (you can send it in private if you want)
Jun 13 23:41:36 <mr_claus> the problem is that i dont can login to yahoo
Jun 13 23:42:13 <ka6sox-away> create one.
Jun 13 23:42:26 <ka6sox-away> VirtualCtor, what page did you ask for it on?
Jun 13 23:42:28 <mr_claus> i log in and get the message "take a look at your mail account, you will get a mail from Yahoo Services"
Jun 13 23:42:59 <VirtualCtor> Um, the " When you wish upon a slug..." page.
Jun 13 23:43:10 <ka6sox-away> nm...I see it.
Jun 13 23:43:27 <ka6sox-away> raise the point in the ML also.
Jun 13 23:43:48 <ka6sox-away> it will get better exposure there too.
Jun 13 23:43:55 <ka6sox-away> (and can be discussed)
Jun 13 23:43:56 <VirtualCtor> Okay, I'll mention it.
Jun 13 23:43:56 <mr_claus> arggll, why not a simple mailman :/
**** ENDING LOGGING AT Mon Jun 13 23:59:56 2005