[12:08] <jbailey> BenC: He's pretty certainly asleep right now.
[12:08] <jbailey> He usually starts his workday in 6 hours.
[12:09] <BenC> tx
[03:38] <zul> BenC: ping
[03:53] <zul> i know you are online at least 40 minutes ago ;)
[03:54] <zul> so you cant hide forever
[04:01] <BenC> here
[04:02] <zul> hi...so i did a clone and did a checkout but im getting all these errors about sha1 and the like
[04:03] <zul> i did git checkout ubuntu-2.6.14 in the directory that was created by the clone so im not sure
[04:28] <zul> meh...stupid git im off to bed
[04:30] <zul> ttyl
[05:56] <fabbione> BenC: pong?
[05:56] <fabbione> morning
[02:22] <BenC> fabbione? 
[02:53] <fabbione> BenC: ?
[03:00] <BenC> fabbione: cluster 1.01.00 code is in git now
[03:01] <BenC> fabbione: also, you need to "git checkout -f master" and "git pull" to get out of the branch and back to where I am working
[03:01] <BenC> working on branches is a bit of a pita with git it seems
[03:01] <BenC> it doesn't handle things for you so easily as working from the head
[03:01] <fabbione> BenC: ah ok
[03:02] <BenC> pull from rsync.kernel.org too
[03:02] <BenC> seems rookery is broken
[03:02] <fabbione> BenC: ok .. i guess i will need to read a full doc
[03:02] <fabbione> roockery has been fixed a few hours ago 
[03:02] <BenC> ah, well, I haven't pushed to it yet, so r.k.o is where all the current code is :)
[03:03] <fabbione> BenC: if we need to kill the branch i guess i will have to redo everything
[03:03] <fabbione> because i did create my archive on roockery and pulled from it
[03:04] <fabbione> and i agree that working from branches is a pain
[03:04] <BenC> do you have lots of local commits?
[03:04] <BenC> you can pull from the branch to get them back
[03:04] <fabbione> for the cluster stuff, did you take it from CVS or from the release?
[03:04] <BenC> release
[03:04] <fabbione> kill it
[03:04] <fabbione> we need to pull from CVS
[03:04] <fabbione> no i have no local commits at all
[03:04] <BenC> wouldn't it be easier to just update it from CVS?
[03:05] <fabbione> no, because the release is for redhat with redhat bits only in
[03:05] <BenC> ok
[03:08] <fabbione> BenC: does the actual kernel builds?
[03:08] <fabbione> or are you just pulling pieces together?
[03:09] <BenC> it all builds now
[03:09] <fabbione> ok
[03:09] <BenC> still getting some small patches from breezy into it
[03:09] <fabbione> ok
[03:09] <BenC> major drivers and patches are applied though
[03:10] <fabbione> yup.. just leave alone the cluster and ocfs2 stuff
[03:10] <fabbione> i will take of them as soon as we stabilize the push/pull thingy
[03:10] <BenC> some of hte questionable "good enough for breezy to release, but not good enoug to submit" patches, I am reviewing
[03:10] <fabbione> because they must be synced with userland
[03:10] <BenC> ok
[03:10] <fabbione> it's pulling 289 objects
[03:11] <BenC> yeah, there's a lot
[03:11] <fabbione> yeps
[03:11] <fabbione> stay tuned that we can test a full push/pull
[03:12] <BenC> ok
[03:14] <fabbione> i don't see any ubuntu commits if i do the checkout -f master
[03:16] <Lathiat> hrm, what are the problems ubuntu has with gcc4 for the kernel?
[03:16] <Lathiat> suse appear to use gcc4 and i hear of others doing it
[03:16] <BenC> fabbione: git pull rsync://rsync.kernel.org/...
[03:17] <fabbione> i assume i need to skip the branch name at the end
[03:17] <BenC> correct
[03:17] <fabbione> ok
[03:17] <BenC> hoping to get the ubuntu-git-guide on wiki today
[03:18] <fabbione> Lathiat: not all the drivers we have comppiles with gcc-4.0. gcc4 miscompile the kernel. patches are welcome. kthxbye :)
[03:18] <fabbione> BenC: yes.. that'd be really nive
[03:18] <fabbione> nice
[03:18] <dereks__> anyone successfully get amd64-smp up?
[03:19] <fabbione> dereks__: yes.
[03:19] <fabbione> BenC: pulling now
[03:19] <BenC> fabbione: I need to go smoke a cig and drink some coffee
[03:19] <dereks__> fabbione: alright, must be my machine. i will keep working on it
[03:19] <BenC> be back in 10 minutes
[03:19] <fabbione> BenC: i need a cig too :) didn't know you smoke ;)
[03:19] <fabbione> brb
[03:20] <BenC> fabbione: we should share room at UBZ then :)
[03:20] <fabbione> BenC: i am with kiko already :/
[03:20] <fabbione> because i will be there all 2 and 1/2 weeks :(
[03:20] <fabbione> so they arranged room mates by arrival/departure
[03:20] <fabbione> BenC: next conf we will
[03:23] <Lathiat> fabbione: so is it more an external driver thing or? (i mean, i heard it miscompild the kernel, i was just curious why other people seem to be happily using it)
[03:29] <fabbione> Lathiat: it's both
[03:45] <fabbione> meeeeehe it's pulling down an entire .pack
[03:46] <fabbione> this is so messy
[03:49] <BenC> yeah, I did a pack of my changes
[03:49] <fabbione> yes, but it's pulling the one from linux
[03:49] <fabbione> linus
[03:49] <BenC> well, that's probably a good thing
[03:50] <fabbione> BenC: the point is that i find it all very confusing
[03:50] <fabbione> i sort of lost track of what is pulling and why
[03:50] <fabbione> i did the local clone on roockery
[03:50] <fabbione> created fabbione-2.6.16.git
[03:50] <fabbione> cloned locally
[03:51] <fabbione> now i am trying to get the local one to be in sync
[03:51] <fabbione> and it's pulling world from everywhere
[03:51] <fabbione> there was no .pack from kernel.org when pulling from your archive
[03:52] <fabbione> BenC: should we get roockery in sync and see step by step what needs to be done?
[03:52] <fabbione> and write it directly on the wiki?
[03:54] <jbailey> Is rookery unfucked now?
[03:54] <fabbione> jbailey: yes.. Znarl did fix it a few hours back
[03:54] <jbailey> Cool.
[04:00] <BenC> fabbione: the repo on rookery you have is just so I can pull from
[04:01] <BenC> fabbione: technically, you should pull from rsync.k.o, and push to rookery for me
[04:01] <fabbione> BenC: ah ok
[04:01] <fabbione> but in theory i can just pull from your repo on roockery.. it should make no difference from kernel.org, right?
[04:01] <BenC> my repo on rookery is just to make it easier (and smaller) for you and others to create a repo I can pull from
[04:02] <fabbione> BenC: ahhhhh ok
[04:02] <BenC> correct, except that the repo on r.k.o is sharing objects with Linus' tree, much the same way that your rookery repo is sharing objects with my repo on rookery
[04:02] <BenC> the .pack thing is probably caused by the fact that my rookery repo has the .pack unpacked
[04:03] <fabbione> gotcha
[04:04] <fabbione> i guess this is going to take the usual 213 hours of rsync
[04:05] <BenC> the second (larger) .pack is about 70Megs
[04:06] <fabbione> it's already stalling
[04:07] <fabbione> this network is the sucks
[04:08] <fabbione> 7.6M now
[04:08] <fabbione> it will take hours
[04:12] <fabbione> bah fuck.. rsync is dead
[04:12] <Mithrandir> good thing you're using rsync so you can restart, then. :-P
[04:12] <fabbione> Mithrandir: not with git...
[04:12] <Mithrandir> ew, ok
[04:12] <fabbione> anywa.. bbl
[04:13] <fabbione> swithing network
[04:13] <fabbione> at least i can work
[04:42] <fabbione> BenC: still around?
[05:12] <BenC> yeah
[05:13] <fabbione> BenC: ok pulling now on a decent VLAN
[05:13] <fabbione> i am at the second pack already
[05:13] <fabbione> should be done in a few minutes
[05:16] <BenC> ok
[05:26] <fabbione> BenC: ok.. i am all set with a clean checkout now
[05:26] <fabbione> let's try again the all commit/push dance :)
[05:28] <fabbione> ok commit done..
[05:28] <fabbione> let's see if i can push to roockery
[05:30] <fabbione> BenC: ok.. it seems to work
[05:30] <fabbione> you should be able to pull from http://people.ubuntu.com/~fabbione/archives/fabbione-2.6.14.git or the ssh equivalent
[05:31] <fabbione> it's the same fix i did last time
[05:31] <fabbione> so nothing fancy
[05:31] <fabbione> just a test commit
[05:34] <fabbione> BenC: i just noticed that to fix the cloop header you also commited a USB driver :P
[05:38] <BenC> really?
[05:39] <BenC> that can't be right
[05:39] <BenC> that driver was already there
[05:39] <fabbione> gitk says otherwise
[05:40] <fabbione>  [UBUNTU:block/cloop]  Fix include of header to be cloop.h
[05:40] <fabbione> in this commit
[05:40] <BenC> yeah, I see it
[05:41] <BenC> but it shouldn't be there
[05:41] <BenC> oh, that was just a file rename
[05:41] <BenC> accident
[05:41] <fabbione> ehhe ok :)
[05:42] <fabbione> anyway can you try to pull from me?
[05:42] <BenC> yeah
[05:42] <fabbione> just to see if it works as it should
[05:42] <fabbione> and close the circle
[05:42] <fabbione> like push to kernel.org
[05:42] <fabbione> and i update from kernel.org
[05:42] <fabbione> if all of the above works
[05:42] <fabbione> we are setted :)
[05:43] <BenC> commit f5608da0ec39df73f62ec752c356fd41ef159150
[05:43] <BenC> Author: Fabio M. Di Nitto <fabbione@ubuntu.com>
[05:43] <BenC> Date:   Fri Oct 21 17:27:34 2005 +0200
[05:43] <BenC>     [UBUNTU/debian] : Test commit - Fix Ben email address
[05:43] <BenC>     Signed-off-by: Fabio M. Di Nitto <fabbione@ubuntu.com>
[05:43] <BenC> looks like we are good
[05:43] <fabbione> did you push to kernel.org already?
[05:43] <BenC> ...
[05:44] <BenC> done to kernel.org
[05:45] <BenC> [UBUNTU:scsi]  Restore generic SCSI proc_info function
[05:46] <BenC> that was my last commit
[05:46] <fabbione> hmm
[05:47] <fabbione> the last one i see is the [UBUNTU:fs/gfs]  Kconfig cleanup
[05:47] <fabbione> same as on kernel.org git web
[05:47] <BenC> not sure of the delay between master.kernel.org and rsync.kernel.org
[05:47] <mkrufky> its horrible
[05:47] <mkrufky> they're working on it
[05:47] <mkrufky> :-(
[05:47] <fabbione> ah ok
[05:48] <mkrufky> are you guys going to use gitweb also?  i find it EXTREMELY helpful
[05:48] <fabbione> mkrufky: horrible to what degree?
[05:48] <fabbione> mkrufky: gitk > gitweb
[05:48] <BenC> how horrible is horrible?
[05:48] <BenC> fabbione: s/rsync.kernel.org/master.kernel.org
[05:48] <mkrufky> fabbione: well, linux released 2.6.14-rc4 at about 3am yesterday (my time) and i didnt see it on kernel.org until after 8pm
[05:48] <fabbione> mkrufky: given that we push to kernel.org i don't really think we need to setup one...
[05:48] <mkrufky> s/linux/linus
[05:48] <fabbione> BenC: yeah
[05:48] <mkrufky> fabbione: good point
[05:49] <BenC> see if it lets you pull from there
[05:49] <mkrufky> i want v4l/dvb to start using git... but there is MUCH resistance
[05:49] <fabbione> BenC: i think rsync on master is open only for users
[05:51] <BenC> possibly
[05:51] <fabbione> yeah
[05:52] <BenC> fabbione: rookery is almost in sync, so in a minute you can try a pull from there
[05:52] <fabbione> BenC: sure...
[05:52] <BenC> hmm, master.k.o and rsync.k.o show the same IP address
[05:52] <fabbione> now it's pulling stuff
[05:53] <BenC> no wait
[05:53] <BenC> they aren't the same IP
[05:53] <BenC> rookery is up-to-date
[05:53] <fabbione> so is rsync kernel.org
[05:54] <fabbione> commit 721ee6203e8989543dd60127c045bbf29c644fc8
[05:54] <fabbione> Merge: 192081598952aa399e68608449a81aa44edddd06 f5608da0ec39df73f62ec752c356fd41ef159150
[05:54] <fabbione> Author: Ben Collins <bcollins@ubuntu.com>
[05:54] <fabbione> Date:   Fri Oct 21 11:41:00 2005 -0400
[05:54] <fabbione>     Merge ssh://people.ubuntu.com/home/fabbione/public_html/archives/fabbione-2.6.14
[05:54] <BenC> cool, that was only a few minutes
[05:54] <fabbione> and later there is scsi_proc
[05:54] <fabbione> yeps
[05:54] <BenC> so our develpment circle is complete
[05:55] <fabbione> BenC: yes
[05:55] <fabbione> we need to decide a couple of things
[05:55] <fabbione> but i guess we can discuss them at UBZ
[05:56] <fabbione> like we want to play tag team
[05:56] <fabbione> or how do we release if one of us (specially you) isn't around
[05:56] <BenC> btw, my commit templates are in debian/commit-templates
[05:57] <fabbione> ah nice
[05:57] <BenC> -s is a requirement to a commit
[05:57] <BenC> yeah, I've been wondering about that myself
[05:58] <fabbione> yes.. my main concern is more political than technical
[05:58] <fabbione> in terms that we should push to kernel.org automatically
[05:58] <fabbione> if possible at all
[05:58] <fabbione> and you should be able to pull from me automatically
[05:58] <fabbione> in terms to keep our trees always in sync
[05:59] <fabbione> so that if i need to upload
[05:59] <fabbione> and you are not around
[05:59] <fabbione> i will upload the same tree that's on kernel.org or roockery
[05:59] <fabbione> we don't want people to say: Hey they upload before pushing patches around
[05:59] <fabbione> or stuff like that
[05:59] <fabbione> it would kill our team image
[06:01] <BenC> it wont build from debian/rules
[06:01] <fabbione> oh
[06:01] <BenC> just copy or create a config
[06:01] <fabbione> ah ok
[06:02] <BenC> none of the configs are updated yet, since we still have some drivers to put in place
[06:02] <fabbione> BenC: yes, fair enough
[06:02] <fabbione> BenC: if you want on monday i can start cleaning up debian/rules
[06:03] <BenC> yeah, that'd be nice
[06:03] <fabbione> even if we don't update configs 100% we can still get the system to build
[06:03] <fabbione> i usually did like that
[06:03] <BenC> I'm doing configs today
[06:03] <fabbione> using old configs to test build
[06:03] <fabbione> i am pretty sure it will fail on the cluster stuff :/
[06:03] <fabbione> the code is for kernel < 2.6.12
[06:03] <fabbione> so just disable it in the config
[06:03] <BenC> nah, the cluster stuff built without even a warning
[06:03] <fabbione> it did?????
[06:04] <fabbione> oh right
[06:04] <fabbione> it's the ABI changes that will make it crash later :)
[06:04] <BenC> yeah, dlm, cman, gfs and all gfs_locking modules
[06:04] <fabbione> ABI/API
[06:04] <fabbione> we also need to think about hppa
[06:04] <fabbione> because it's the only arch that requires a specific patch to build
[06:04] <fabbione> they just created a git repo
[06:04] <BenC> I think hppa might be the only thing that keeps dpatch around
[06:04] <fabbione> and pushing to linus
[06:05] <fabbione> for one patch we can apply/deapply manual
[06:05] <fabbione> i wouldn't even bother to keep dpatch around for that
[06:05] <BenC> true
[06:06] <BenC> how did you create the flavor configs? Did you start with one config, and change things for say 686, 686-smp, k7, k7-smp, etc?
[06:06] <fabbione> BenC: anyway another little hint is to update the configs directly when building the first pkg set
[06:06] <BenC> and id so, is it documented what the exact config options should be for 686, etc?
[06:06] <fabbione> BenC: i didn't start them.. Xu did
[06:06] <fabbione> i only updated them in time
[06:06] <fabbione> i think it would be better to scratch them
[06:06] <fabbione> and redo all of them from scratch
[06:06] <BenC> I did a diff and they seem to be a little out of sync
[06:06] <fabbione> that's why
[06:07] <fabbione> it was one of the thing to do for breezy
[06:07] <fabbione> -ENOTIME
[06:07] <BenC> I've been thinking about a new config system for us
[06:07] <fabbione> BenC: did you look at the debian config system?
[06:07] <BenC> keep all the common config options (per arch, and global) and tack on flavor specific options at build
[06:07] <fabbione> they did implement something that seems to be nice based on our idea
[06:07] <BenC> no
[06:07] <fabbione> you just said exactly what they did implement
[06:07] <fabbione> and that was coming from UDU bof ;)
[06:07] <BenC> I'll check it out then
[06:07] <fabbione> yeah
[06:07] <fabbione> it's not worth to rewrite
[06:08] <fabbione> and we need to push the arches in sync
[06:08] <fabbione> something that we don't really do
[06:08] <BenC> we're a little out of sync even in the same arch, between flavors, but yeah, between archs we are even worse
[06:08] <fabbione> so if it's ok with you, i will start cleaning debian/* to be able to build, kill useless crap
[06:09] <fabbione> and try to make it nicer
[06:09] <BenC> sure
[06:09] <fabbione> (that's from monday)
[06:09] <fabbione> it's kind of we here already :)
[06:09] <fabbione> i will leave you the config management stuff
[06:10] <fabbione> also remember that some of the extra features we added like cluster
[06:10] <fabbione> can be temporary disabled while we stabilize the build system
[06:10] <fabbione> it's a development kernel.. so we can readd things later
[06:13] <fabbione> BenC: dude.. have a nice weekend
[06:13] <fabbione> cya on monday 
[06:30] <jbailey> BenC: Still around?
[06:39] <mkrufky-away> jbailey: sorry to see-saw on the dvb-headers issue -- it is NOT fixed ... we're gonna need those available, if possible
[06:39] <mkrufky-away> (if you want ubuntu users to be able to compile new dvb modules from cvs)
[06:39] <jbailey> mkrufky-away: It's too late for breezy now.
[06:39] <mkrufky-away> ya thats ok... .im saying for dapper
[06:40] <jbailey> mkrufky-away: 'k.  Won't it be fixed by moving the headers over anyway>
[06:40] <BenC> jbailey: yeah
[06:40] <mkrufky-away> dvb maintainer has been hard to reach lately ... he knows about the issue, and he will move them to /include/media/dvb ... it's on his to-do list
[06:40] <mkrufky-away> jbailey: why not?  (u say wont be fixed)
[06:41] <jbailey> BenC: drow was asking me if we have a wiki page or something that covers how to do custom kernels.
[06:41] <jbailey> BenC: Got anything I can give him?
[06:41] <BenC> custom from user provided source, or our source?
[06:42] <mkrufky-away> jbailey: this is the best howto around :
[06:42] <mkrufky-away> http://www.us.debian.org/releases/stable/i386/ch08s05.html.en#id2527454
[06:42] <jbailey> I think he wants our source.  Apparently there's a trackpad driver tht he needs for a thinkpad.
[06:42] <jbailey> mkrufky-away: I don't know that our kernels look much like Debian's anymore.
[06:42] <mkrufky-away> doesnt matter
[06:42] <mkrufky-away> compilation still will work
[06:43] <BenC> jbailey: wiki.ubuntu.com/CategoryKernel
[06:43] <BenC> there's a few links in there
[06:45] <jbailey> BenC: Luvly, thanks.
[07:44] <zul> BenC: i upgraded my version of git and it works now for some reason :)
[09:20] <BenC> zul: must be that it needed to understand alternatives
[09:22] <zul> true