**** BEGIN LOGGING AT Sun Aug 07 23:59:56 2005
**** BEGIN LOGGING AT Mon Aug 08 00:57:19 2005
**** BEGIN LOGGING AT Mon Aug 08 01:05:35 2005
**** BEGIN LOGGING AT Mon Aug 08 01:28:04 2005
Aug 08 02:44:56 <NAiL__>  /nick NAiL
Aug 08 02:45:06 <NAiL__> duh
Aug 08 03:42:19 <odoc> Erase from 0x50000000-0x50800000: ................................................................
Aug 08 03:42:19 <odoc> Err = 9910
Aug 08 03:42:19 <odoc> Can't program region at 0x50000000: Error trying to program
Aug 08 03:42:21 <odoc> ?
Aug 08 03:43:04 <rwhitby> sure you've got the right numbers there?
Aug 08 03:43:48 <odoc> got it from the wiki
Aug 08 03:43:54 <odoc> want to erase everything 
Aug 08 03:44:05 <odoc> including cfe
Aug 08 03:44:08 <odoc> my nslu is weird
Aug 08 03:44:25 <odoc> now without cfe:
Aug 08 03:44:27 <odoc> Can't program region at 0x50060000: Error trying to program
Aug 08 03:44:28 <rwhitby> cfe?
Aug 08 03:44:28 <odoc> ?
Aug 08 03:44:39 <rwhitby> do you mean redboot?
Aug 08 03:45:36 <odoc> aehh
Aug 08 03:45:38 <odoc> yes
Aug 08 03:45:45 <odoc> openwrt ;P
Aug 08 03:45:59 <rwhitby> so you're trying to erase redboot whilst running redboot?
Aug 08 03:46:20 <rwhitby> I sure hope you have confirmed JTAG access on that slug ..
Aug 08 03:46:43 <odoc> jtag shouldnt be a problem
Aug 08 03:47:11 <rwhitby> shouldn't be, or isn't ?
**** BEGIN LOGGING AT Mon Aug 08 03:47:49 2005
Aug 08 03:48:32 <odoc> its not working probably anyway
Aug 08 03:48:34 <odoc> and i have wrt-jtag experience ;P
Aug 08 03:49:15 <odoc> odoc its not working probably anyway
Aug 08 03:49:18 <odoc> odoc and i have wrt-jtag experience ;P
Aug 08 05:30:18 <CIA-7> 03marceln * 10unslung/ (4 files in 4 dirs): Moved chown nobody to postinst
Aug 08 06:07:46 <rwhitby> I've updated the wiki sidebar a bit ..
Aug 08 08:52:42 <Crofton> um, I just booted my slug and am wondering what the default pw is?
Aug 08 08:55:02 <Tiersten> uNSLUng
Aug 08 08:55:44 <Tiersten> it's in the FAQ
Aug 08 08:58:49 <Crofton> actually, I needed the linsys default, but it is also in the FAW
Aug 08 08:58:52 <Crofton> er FAQ
Aug 08 08:58:54 <Crofton> thanks
Aug 08 08:59:06 <Tiersten> We don't know the Linksys default
Aug 08 08:59:33 <Crofton> well, the faq answered the question
Aug 08 08:59:42 <Tiersten> Considering the generated ones for ourtelnetrescueuser look like 543oijioj5io234j5o23 then it is probably something equally obscure
Aug 08 08:59:53 <Crofton> I don;t know it either because I left the docs at home
Aug 08 09:00:45 <Crofton> Hopefully I never have to worry about it again, looking at building my own firmware image now
Aug 08 09:01:12 <Tiersten> docs at home?
Aug 08 09:01:29 <Tiersten> Nobody knows the root password apart from Linksys/Sercomm
Aug 08 09:01:46 <Crofton> oh, I meant for the admin interface
Aug 08 09:01:50 <Tiersten> Ohhh
Aug 08 09:02:01 <Tiersten> Okay.  We know that :)
Aug 08 09:02:10 <Crofton> I was confused for a bit
Aug 08 09:02:32 <Crofton> I hadn't thought about the case were it must have a real root pw also
Aug 08 09:02:47 <Crofton> I had the cover off even before I powered it up
Aug 08 09:03:15 <Tiersten> heh
Aug 08 09:04:07 <Tiersten> Linksys aren't too afraid about people opening the NSLU2 at the moment.  The WRT54G(S) have warranty stickers over the casing to stop you from opening it
Aug 08 09:04:18 <Crofton> A friend has a digital cam with a busted LCD
Aug 08 09:04:25 <Crofton> heh
Aug 08 09:04:28 <Crofton> warrranty 
Aug 08 09:04:31 <Crofton> what is that?
Aug 08 09:04:59 <Crofton> Hopefully I can get wireless and gphoto in the 8m of flash :)
Aug 08 09:06:04 <Tiersten> You should be able to...
Aug 08 09:06:16 <Tiersten> Not th eend of othe world if you can't anyway.  Buy a cheap USB memory stick
Aug 08 09:06:38 <Crofton> trying to avoid needing a usb hub
Aug 08 09:06:47 <Tiersten> ah
Aug 08 09:06:52 <Crofton> since we want wireless also
Aug 08 09:06:59 <Crofton> it is a hobby :)
Aug 08 09:07:26 <Tiersten> What wireless dongle you going to use?
Aug 08 09:07:31 <Crofton> not sure
Aug 08 09:07:38 <Crofton> I have one or two lying around
Aug 08 09:07:41 <Crofton> cheap ones
Aug 08 09:07:59 <Crofton> I had it working on a laptop at one point
Aug 08 09:08:10 <Tiersten> No binary modules tho?
Aug 08 09:08:11 <Crofton> usb wireless drivers are bitchy to build
Aug 08 09:08:16 <Crofton> no
Aug 08 09:08:24 <Tiersten> k.  Just checking
Aug 08 09:08:45 <Crofton> heh, ndiswrapper on a slug would be "amusing"
Aug 08 09:09:02 <Tiersten> Some people are trying to get a zd1211 based USB dongle working.  There are open sourced drivers for it so it should work
Aug 08 09:09:21 <Tiersten> I've got a couple RT2500 based ones as well which have open sourced drivers but I've not tried it yet
Aug 08 09:09:22 <Crofton> I need to check on the ones I have
Aug 08 09:09:31 <Crofton> they are older ones
Aug 08 09:09:40 <Crofton> one is prism I think
Aug 08 09:10:45 <Tiersten> b or g?
Aug 08 09:10:50 <Crofton> b
Aug 08 09:11:40 <Tiersten> If it's a Prism 2 based one then it should work...
Aug 08 09:11:47 <Tiersten> Looks a bit of a hassle to compile as you said though
Aug 08 09:11:55 <Crofton> yep
Aug 08 09:12:17 <blaster8> the RT2500 drivers are a mess
Aug 08 09:12:22 <Crofton> seems like wireless driver writing attracts people who have trouble getting along with others
Aug 08 09:12:33 <blaster8> but in Linux no-one cares so no-one changes them
Aug 08 09:12:35 <Tiersten> blaster8: Which ones?
Aug 08 09:12:45 <blaster8> the ones that Ralink wrote
Aug 08 09:12:51 <Tiersten> Ah okay
Aug 08 09:12:56 <Tiersten> There is an open source project as well
Aug 08 09:13:10 <Tiersten> The Ralink ones look like the Windows ones but with some shims to make it work with Linux
Aug 08 09:13:19 <blaster8> OpenBSD just rewrote the drivers from scratch in 1/10th of the space
Aug 08 09:13:37 <blaster8> http://kerneltrap.org/node/4818
Aug 08 09:14:23 <blaster8> as well as rewriting the Atheros 802.11g HAL layer, where Linux has to use the proprietary extension
Aug 08 09:15:03 <Tiersten> Cool
Aug 08 09:15:28 <blaster8> indeed - I do think the who Linux distribution thing doesn't really help
Aug 08 09:15:55 <blaster8> in BSD it seems that stuff gets integrated much better - but unfortunately Linux has the more advanced kernel
Aug 08 09:30:32 <kolla_> more messy you mean? :)
Aug 08 09:31:28 <blaster8> well, definitely more messy - but a kernel that can fork with practically no overhead, for example is pretty cool
Aug 08 09:31:53 <Tiersten> What?  clone?
Aug 08 09:33:17 <BOfH> umount: /media/hdd: Device or resource busy
Aug 08 09:33:24 <BOfH> Keep getting that error
Aug 08 09:33:43 <BOfH> /dev/sdb1 is mounted there, it's an external drive, I just yanked the USB plug
Aug 08 09:33:48 <Tiersten> Stop the processes that are still on there
Aug 08 09:33:51 <Tiersten> You shouldn't yank the plug
Aug 08 09:34:28 <BOfH> Only thing using it is [scsi_eh_2] and [scsi_eh_3]
Aug 08 09:34:42 <Tiersten> You ran lsof?
Aug 08 09:34:50 <Tiersten> Those two are kernel threads anyway
Aug 08 09:34:55 <BOfH> command not found
Aug 08 09:35:05 <BOfH> yea, but they're keeping /dev/sdb busy
Aug 08 09:35:11 <BOfH> When the device has been removed
Aug 08 09:35:27 <Tiersten> They're not keeping it busy... they're the actual parts that handle it...
Aug 08 09:35:33 <BOfH> Ah
Aug 08 09:35:38 <Tiersten> Something else was still in that directory which is why you couldn't unmount
Aug 08 09:35:43 <BOfH> Well, they're "handling" a nonpresent device
Aug 08 09:35:57 <Tiersten> It could be that you're running something off it or something has it's current working directory still inside there
Aug 08 09:36:11 <Tiersten> You're not supposed to yank the plug.  That's why it's not liking it
Aug 08 09:36:21 <BOfH> Then what are you supposed to do?
Aug 08 09:36:33 <Tiersten> Stop whatever process is still using it?
Aug 08 09:36:38 <Tiersten> At this point, reboot it
Aug 08 09:36:42 <BOfH> They were all stooped?
Aug 08 09:36:46 <BOfH> stopped*
Aug 08 09:36:57 <Tiersten> No.  Something was still in there otherwise it wouldn't complain
Aug 08 09:37:05 <Tiersten> Are you sure you're not in that directory in your shell?
Aug 08 09:37:28 <BOfH> ...
Aug 08 09:37:29 <BOfH> duh
Aug 08 09:37:35 <Tiersten> ?
Aug 08 09:37:45 <BOfH> I was in that dir in another ssh session
Aug 08 09:37:49 <Tiersten> ah
Aug 08 09:38:00 <Tiersten> Anyway, if it's still thinking it's there then the easiest way to fix it is to reboot
Aug 08 09:38:13 <BOfH> Can't stop the kernel proc any other way?
Aug 08 09:38:24 <Tiersten> No
Aug 08 10:10:50 <kolla_> unable to umount /proc?
Aug 08 12:52:10 <jp30-work> marceln, ping?
Aug 08 13:54:45 <jp30-work> marceln, i ping you again
Aug 08 13:57:50 <CIA-7> 03jp30 * 10unslung/make/openssl.mk: stage openssl.pc
Aug 08 14:34:13 <marceln> Sorry?
Aug 08 15:23:21 <dyoung-zzzz> test
Aug 08 15:23:56 <rwhitby-away> 1 2 3
Aug 08 15:24:29 <jp30-work> crrk... turn the foldback up please...
Aug 08 15:26:35 <CIA-7> 03jp30 * 10unslung/make/file.mk: depend on zlib, strip, and upstream update to 4.13
Aug 08 15:37:08 <rwhitby-away> jp30-work: are the autobuilds working out for you?  Are you able to trust it to always build properly for you automatically, and is that saving you time?
Aug 08 15:38:16 <rwhitby-away> (note that we've got optware/unslung native autobuilding now too)
Aug 08 15:41:34 <jp30-work> rwhitby, yes, they're saving me time. it's great. just one issue...
Aug 08 15:41:36 <dyoung-zzzz> morning gernika 
Aug 08 15:41:47 <gernika> evening dyoung
Aug 08 15:42:35 <jp30-work> i'd like to be able to trigger the autobuilder to dirclean and rebuild a package without bumping the ipk version. (for example, if the staging procedure has changed for a package, but not the ipk contents)
Aug 08 15:42:42 <gernika> Looks like logging is working
Aug 08 15:43:58 <dyoung-zzzz> sorry, I guess it is.
Aug 08 15:44:12 <dyoung-zzzz> I need to learn to type in the morning apparently. :-)
Aug 08 15:44:52 <gernika> Well, for a while, all the messages were being timestamped a day behind, so that may have been confusing.  I just fixed that about 45 min. ago.
Aug 08 15:45:45 <dyoung-zzzz> okay cool, thanks!
Aug 08 15:46:07 <rwhitby-away> dyoung-zzzz: do we have -general and opeslug channels covered now too?
Aug 08 15:46:37 <rwhitby-away> gernika: thx again for providing us with a real-time logging capability
Aug 08 15:47:13 <rwhitby> jp30-work: do you have a proposal on how to trigger that?
Aug 08 15:47:56 <jp30-work> hmm... could make autoclean more aggressive, so it cleans any package older than its own .mk file
Aug 08 15:48:11 <rwhitby> jp30-work: that would work for me
Aug 08 15:48:39 <gernika> rwhitby: you're welcome
Aug 08 15:48:54 <jp30-work> ok, and i can do that without your intervention too, as it's just changes to optware
Aug 08 15:48:59 <rwhitby> yep
Aug 08 15:49:09 <jp30-work> next time i need this, i'll hack it in
Aug 08 15:49:17 <rwhitby> that'll work for both cross and native too
Aug 08 15:49:48 <jp30-work> i'm interested in some feedback on a grand idea for optware...
Aug 08 15:50:13 <jp30-work> i hacked together a preprocessor for optware .mk files a couple weeks ago
Aug 08 15:50:44 <jp30-work> it allows abstracting the content of template.mk away from the contents of the .mk files
Aug 08 15:51:04 <jp30-work> will show an example...
Aug 08 15:52:02 <jp30-work> ok, here's the contents of "xproto.mk.in" as an example:
Aug 08 15:52:03 <jp30-work> PACKAGE=xproto
Aug 08 15:52:03 <jp30-work> SITE=http://freedesktop.org/
Aug 08 15:52:03 <jp30-work> REPOSITORY=:pserver:anoncvs@freedesktop.org:/cvs/xlibs
Aug 08 15:52:05 <jp30-work> DIR=Xproto
Aug 08 15:52:07 <jp30-work> CVS_DATE=20050130
Aug 08 15:52:09 <jp30-work> MAINTAINER=Josh Parsons <jbparsons@ucdavis.edu>
Aug 08 15:52:13 <jp30-work> DESCRIPTION=X protocol headers
Aug 08 15:52:15 <jp30-work> SECTION=lib
Aug 08 15:52:17 <jp30-work> PRIORITY=optional
Aug 08 15:52:19 <jp30-work> USES=cvs automake
Aug 08 15:53:15 <jp30-work> the preprocessor puts this together with the contents of "templates/cvs.mk" "templates/automake.mk" to make the same content that's currently in xproto.mk
Aug 08 15:53:24 <jp30-work> what do people think of this idea?
Aug 08 15:57:31 <rwhitby> interesting ....
Aug 08 15:57:33 <dyoung-zzzz> rwhitby: it should in theory all work now.
Aug 08 15:58:01 <rwhitby> jp30-work: how would we handle the difficult one-off cases?
Aug 08 15:59:32 <jp30-work> rwhitby, sorry phone just rang... with you in 15min
Aug 08 15:59:38 <rwhitby> ok
Aug 08 16:12:12 <jp30-work> ok, i'm back...
Aug 08 16:13:33 <jp30-work> the preprocessor actually combines both makefile definitions and rules. so, in the tricky cases, you would have a rule in the package .mk file, and the preprocessor would know not to copy that rule from any of the templates
Aug 08 16:14:09 <jp30-work> i can show an example of that as well
Aug 08 16:16:16 <jp30-work> here's a fragment of crosstool.mk, showing the syntax of an overridden rule:
Aug 08 16:16:24 <jp30-work> PACKAGE=crosstool
Aug 08 16:16:24 <jp30-work> VERSION=$(_CROSSTOOL_VERSION)
Aug 08 16:16:24 <jp30-work> DESCRIPTION=GCC and friends
Aug 08 16:16:24 <jp30-work> IPK_VERSION=1
Aug 08 16:16:24 <jp30-work> USES=crosstool
Aug 08 16:16:25 <jp30-work> ### patch for arm
Aug 08 16:16:27 <jp30-work> .unpacked:+
Aug 08 16:16:29 <jp30-work>         mkdir -p $(_BUILD_DIR)/patches/gcc-$(_GCC_VERSION)
Aug 08 16:16:31 <jp30-work>         cp $(_BUILD_DIR)/patches/gcc-3.3.4/gcc-3.3.4-arm-bigendian.patch \
Aug 08 16:16:33 <jp30-work>            $(_BUILD_DIR)/patches/gcc-$(_GCC_VERSION)
Aug 08 16:18:16 <jp30-work> ".unpacked:+" means "add these commands to the $(<FOO>_BUILD_DIR)/.unpacked" rule
Aug 08 16:18:54 <jp30-work> this is a case where a rule is being modified, rather than overwritten
Aug 08 16:20:03 <rwhitby> ok, that looks like a good general solution
Aug 08 16:20:13 <jp30-work> rwhitby, if this meets with your approval i could check it in as a branch to optware cvs
Aug 08 16:21:08 <rwhitby> jp30-work: how will the "add these commands" stuff above handle the case where the last thing a rule does is touch a flag file?
Aug 08 16:21:21 <jp30-work> i'm so pleased you asked me that :)
Aug 08 16:22:13 <jp30-work> here's the definition of .unpacked in templates/base.mk:
Aug 08 16:22:21 <jp30-work> .unpacked: $(_SOURCE_FILE)
Aug 08 16:22:21 <jp30-work>         <<rm -rf $(_BUILD_DIR)
Aug 08 16:22:21 <jp30-work>         <$(MAKE) $(_STAGES_TARGETS)
Aug 08 16:22:21 <jp30-work>         $(_UNZIP) $(_SOURCE_FILE) |tar -C $(BUILD_DIR) -xf -
Aug 08 16:22:21 <jp30-work>         if [ $(BUILD_DIR)/$(_DIR) != $(_BUILD_DIR) ] ; \
Aug 08 16:22:22 <jp30-work>                 then mv $(BUILD_DIR)/$(_DIR) $(_BUILD_DIR) ; \
Aug 08 16:22:24 <jp30-work>         fi
Aug 08 16:22:26 <jp30-work>         >>touch $@
Aug 08 16:22:38 <rwhitby> ok, you've got it covered.  excellent.
Aug 08 16:23:06 <jp30-work> the "<<" and ">>" give commands a priority, and the extra material is added in the middle as a result
Aug 08 16:23:37 <rwhitby> jp30-work: I consider you to be in charge of the optware build system, so as long as autobuilds continue to work (and you can ask me to stop them at any time if absolutely necessary), then feel free to make whatever changes you see fit.
Aug 08 16:33:08 <rwhitby> back later
Aug 08 19:37:03 <malfimalfi> ui, good work at the sidebar 
Aug 08 19:47:00 <rwhitby-web> I added a Downloads page too, since that's what most people look for when they come to a new site :-)
Aug 08 20:39:38 <dyoung-web> ~ubessa
Aug 08 20:39:38 <jbot> well, ubessa is a Unorthrodox But Effective Super Serial Adapter.  It refers to a UBE board using a MAX232 coupled to a 74LVC244 because somebody was too lazy to get a MAX3232.
Aug 08 21:12:45 <CIA-8> 03jp30 * 10unslung/make/ (audiofile.mk esound.mk): fix rpath, deps problems with esound, strip and clean up
Aug 08 21:35:26 <DIrEctQL> hey
Aug 08 21:36:05 <DIrEctQL> how does unslung pivots the / from ramfs to the drive?
Aug 08 21:52:49 <DIrEctQL> anyone?
Aug 08 22:28:43 <CIA-8> 03jp30 * 10unslung/Makefile: promote libmad, madplay - these seemed to have gotten lost earlier
Aug 08 22:55:35 <CIA-8> 03nslu2-linux 07org.openembedded.nslu2-linux * r99c9e931... 10/conf/distro/openslug-packages.conf: Removed softdog from openslug-packages.conf - it don't exist, so why was it added in the first place?
**** ENDING LOGGING AT Mon Aug 08 23:59:57 2005