[00:54] <udsbotu> uds-gb-f: 5 minutes left in this session!
[00:55] <udsbotu> uds-gb-f: 4 minutes left in this session!
[00:56] <udsbotu> uds-gb-f: 3 minutes left in this session!
[00:57] <udsbotu> uds-gb-f: 2 minutes left in this session!
[00:58] <udsbotu> uds-gb-f: 1 minute left in this session!
[00:59] <udsbotu> uds-gb-f: This session has ended.
[14:54] <jamespage> moring xranby
[14:54] <xranby> jamespage: good morning
[14:54] <xranby> di session start in 1h?
[14:54] <xranby> do
[14:55] <jamespage> xranby, yep
[14:55] <jamespage> see you then
[14:55] <xranby> great, yes see you as well have a good breakfast
[15:32] <xranby> gnu_andrew: welcome
[15:35] <gnu_andrew> xranby, thanks for the heads up :-)
[15:37] <xranby> gnu_andrew: http://summit.ubuntu.com/uds-q/meeting/20495/foundations-q-java7/   live notes will get scribbled down here, i think you need a launchpad account to see them
[15:40] <xranby> gnu_andrew here is working at Red Hat and are heavily involved with bringing security fixes into IcedTea from OpenJDK and have done countless IcedTea releases during the last years
[15:42] <nthykier> welcome
[15:46] <gnu_andrew> nthykier, thanks
[15:46] <gnu_andrew> The last date is Oct for OpenJDK 6 security releases
[15:47] <nthykier> 2012?
[15:47] <gnu_andrew> i.e. after the 12 June and 16 Oct releases, there will be no more from Oracle
[15:47] <gnu_andrew> nthykier, yes 2012
[15:47] <nthykier> ugh
[15:47] <gnu_andrew> yeah this is why it's an issue, not just us being nasty :-)
[15:51] <xranby> doko: hi and welcome
[15:54] <jamespage> everyone able to access the pad OK?
[15:56] <nthykier> nope - I guess my membership of that etherpad group needs to be approved first >.>
[15:58] <xranby> ok shall we start?
[15:58] <jamespage> nthykier, I just poked to see if that can be done now
[15:58] <jamespage> yes
[15:59] <jamespage> OK - lets make a start - any late attendees this end can catchup
[16:00] <xranby> for people listening in on audio try reconnect, the shoutcast server restarts every full hour
[16:01] <nthykier> (who is the speaker, btw? jamespage?)
[16:01] <doko> yes
[16:01] <jamespage> https://wiki.ubuntu.com/JavaTeam/Java7Default
[16:03] <gnu_andrew> no access either
[16:04] <xranby> is Ubuntu collaborating with the Fedora team to shake down Java7 FTBFS bugs?
[16:06] <nthykier> A Lintian check for what exactly?
[16:07] <nthykier> (sorry the sound is not the best)
[16:07] <jamespage> nthykier, to check to see if the bytecode in a jar file is only compatible with Java 7
[16:07] <jamespage> i.e. not backwards compatible
[16:07] <doko> nthykier, for the byte-code format used; to make sure that everything is built for e.g. 1.5 or 1.6, not 1.7
[16:07] <nthykier> should be trivial to do - there is some code for that already in javahelper
[16:08] <jamespage> nthykier, are you aware of plans to switch default-java in Debian this release?
[16:09] <nthykier> I am not aware of anyone doing it either, but having heard of the "security support" for Java6, I am considering to work on it
[16:09] <gnu_andrew> nthykier, jamespage: I suggest liasing with Fedora who've already switched in F17.  They may have patches you can share.
[16:10] <doko> usually Debian has a super set of the Fedora packages
[16:10] <gnu_andrew> I know.  But doing the common subset again is silly.
[16:11]  * gnu_andrew remembers only Debian had GNU Classpath and not Fedora :-)
[16:11] <jamespage> agreed - and we will colloborate as much as possible to ensure work is not repeated
[16:12] <gnu_andrew> and yes obviously the ideal is the upstream package has it :-)
[16:14] <xranby> - #ARM ASM-interpreter and Thumb2 JIT ported to #OpenJDK 7u3 Hotspot 22, IcedTea 7 2.1.1 branch, by chrisphi, aph & adinn at Red Hat.
[16:14] <jamespage> gnu_andrew, agreed
[16:14] <xranby> - The only problem is that there  is no IcedTea 7 2.1.1 release planned by the icedtea team. Ubuntu are  currently providing a pre-release build for arm using this 2.1.1 branch.
[16:14] <xranby> - The good news: the assembler port and thumb2 jit work https://docs.google.com/a/gudinna.com/spreadsheet/oimg?key=0ArR4FJuzvXEsdDZkek5QcXd3c2ROMm1IajE1WmFwYkE&oid=4&zx=edub54x8oark
[16:14] <xranby> - all in all my, xranby,  experience is that the openjdk 7 using the arm port   hits no regression  in usage compared to the openjdk 6   port.. so its sane to let  distributions default to it and keep working on tck convergence
[16:14] <gnu_andrew> xranby, plan one :-)
[16:14] <gnu_andrew> xranby, seriously I'd like one but don't feel qualified to say whether we are ready for one or not
[16:16] <xranby> i have asked andrew haley (aph)  Andrew Dinn and Chris Phillips on how much work there is left to get openjdk7  on arm to be feature complete and pass TCK
[16:16] <xranby> (14:03:59) aph_: xranby: I don't think that there are any really specific problems with the ARM port, except for the fact that hotspot is being constantly rewritten.
[16:16] <xranby> (14:04:16) aph_: xranby: apparently permgen is due to be deleted altogether
[16:17] <xranby> (15:50:09) ChrisPhi_:  I see zero  as having a lot of work to converge with jsr 292 / lambda and adopt some bytecode generation approaches, then we should be in maint mode.
[16:18] <jamespage> xranby, thats great news - and agreed that its sane to use in the distro due to no regression against openjdk6 port
[16:18] <xranby> the best part that have happened during the past cycle is that the arm port have passed the tck for openjdk 6
[16:18] <xranby> and that is now actively maintained by 3+ people
[16:20] <xranby> if we want more distributions to use the icedtea7 2.1.1 branch     then i can assign me to a workitem to do one release from that branch in this cycle
[16:20] <xranby> this would create an official icedtea release with the included arm port
[16:20] <doko> xranby, +1
[16:22] <xranby> focus will be heavily switched to java 7
[16:22] <xranby> i think we will drop support on the many parralel icedtea6 branches
[16:25] <xranby> ideally we want icedtea7 2.1.1 branch to pass the tck before releasing it
[16:25] <nthykier> yeah
[16:25] <xranby> are calxeda pleased with the openjdk performance?
[16:25] <doko> xranby, it's the 2.1 branch, 2.1.3 would be fine for this as well
[16:26] <xranby> No
[16:26] <jamespage> xranby, have we got access to TCK for Java7 yet?
[16:26] <jamespage> :-(
[16:26] <xranby> i do not have access to the tck
[16:26] <nthykier> Do we have a "clear" picture of how "far" the Java6 -> Java7 transition is (in terms of FTBFS issues)
[16:26] <xranby> its unclear if i can get access to it since it require signing a nda
[16:26] <xranby> i will work closely on QA with the readhat team
[16:26] <xranby> who have access to the tck
[16:27] <xranby> Red Hat
[16:27] <jamespage> xranby, great - I expect some people in Ubuntu will as well so hopefully we can help there as well
[16:27] <doko> nthykier, jamespage had used some user tags for these, but can't remember which ones ;)
[16:27] <doko> in addition to this, we'll need some gcj patches from trunk for these archs which don't have an openjdk
[16:28] <nthykier> When will the "syncs calm down"?
[16:28] <gnu_andrew> https://fedoraproject.org/wiki/Java7_Package_Rebuild_Status is the Fedora effort if that helps
[16:28] <jamespage> nthykier, ~3weeks ish
[16:29] <xranby> gnu_andrew: thanks
[16:29] <doko> nthykier, they should already, but the merges need to be done
[16:29] <gnu_andrew> doko, we don't even have 2.1.1 yet, steady on :-)
[16:29] <doko> I didn't claim that
[16:29] <jamespage> nthykier, we should have a better view by the end of this month
[16:30] <doko> which maybe is a bit late for the debian freeze
[16:30] <xranby> will ubuntu revert to debian armv4t+ for armel?
[16:30] <nthykier> jamespage: unfortunately "the end of this month" is a little too close to the Debian freezy
[16:30] <doko> xranby, no
[16:30] <nthykier> s/freezy/freeze/
[16:30] <jamespage> nthykier, OK - the implication being we won;t be able to land fixes back into Debian in the short term?
[16:31]  * jamespage is not familiar with the debian release process
[16:31] <gnu_andrew> today is the first I've heard of most of this.  It would be really helpful if we had more input on when/what releases are needed on distro-pkg-dev@openjdk.java.net
[16:31] <nthykier> it will be difficult at best
[16:32] <jamespage> nthykier, esp if they are to support something that is happening in Ubuntu that has no immediate impact in Debian I guess
[16:32] <nthykier> yeah
[16:32] <jamespage> gnu_andrew, what would make life easier?  a head up in advance on what we are doing for the next release as soon as WE know?
[16:33] <nthykier> unless there is a "strong" driving force (e.g. the debian security team) pushing for Java7, I think it will next to impossible to get it in after the freeze starts
[16:33] <jamespage> nthykier, OK - are they aware of the situation with OpenJDK6?
[16:34] <doko> nthykier, unfortunately I didn't see this in the past
[16:34] <nthykier> don't know if they are
[16:34]  * jamespage adds an action
[16:36] <gnu_andrew> jamespage, yeah, and if you need/would prefer a release of some branch
[16:37] <gnu_andrew> doko, nthykier: I doubt Debian wants to be doing their own security support for 6, especially given the limited resources they seem to have on it
[16:37] <xranby> on testing: do we have openjdk tests running on top of the ubuntu   OHIO? test QA framework?
[16:37] <xranby> also .. can we provide preview releases using JuJu to get say openjdk8 running
[16:37] <doko> OHIO?
[16:38] <xranby> doko: whats the QA framework named that the QA team invented during the precise cycle?
[16:38] <gnu_andrew> jamespage, currently I'm mostly shooting in the dark when it comes to releases, which isn't good
[16:38] <jamespage> gnu_andrew, OK - I've actioned myself to talk to you about that...
[16:38] <doko> xranby, the jtreg tests are run with every package upload, same for all third party java packages which provide a testsuite
[16:38] <xranby> jamespage: ideally use juju to compile openjdk 8 from source and install it
[16:39] <jamespage> xranby, possible -= fancy giving it a try?  I can help out
[16:39] <gnu_andrew> jamespage, thanks, much appreciated!
[16:39] <xranby> jamespage: yes i fancy giving it a shot
[16:39] <xranby> to help bring the latest test releases into ubuntu fast
[16:39] <xranby> of openjdk
[16:40] <xranby> i will pass the question to gnu_andrew
[16:40] <xranby> opebnjdk 8 have some interesting new features
[16:40] <xranby> like jigsaw
[16:40] <gnu_andrew> xranby, that's another good point.  I'd like to see 8 moving out earlier so we don't have this issue again.
[16:40] <gnu_andrew> especially as it will change much more than 7 did
[16:42] <gnu_andrew> and getting Debian/Ubuntu's feedback on Jigsaw before it is stable and unchangeable is very important
[16:42] <gnu_andrew> (Jigsaw is the module system for the JDK)
[16:42] <xranby> Jigsaw uses .deb!
[16:42] <xranby> (i think)
[16:43] <nthykier> At least, Debian had a GSoC student on Jigsaw
[16:43] <gnu_andrew> xranby, I've honestly lost track.  dalibor was dabbling with .deb at one point but we really need to get our hands dirty and find out
[16:43] <gnu_andrew> biggest problem at the moment is so much work to do and too few hands on deck to be honest
[16:44] <xranby> i would like to say
[16:44] <xranby> if someone want to help out with a thumb2 -> arm port
[16:44] <xranby> then we can get a jit running on a raspberry pi
[16:45] <xranby> it would be a quite good work for a computer sciense student
[16:45] <xranby> another qusetion
[16:45] <xranby> question: is calxeda in the room?
[16:45] <jamespage> no
[16:46] <xranby> any comments on openjdk performance on their servers?
[16:46] <xranby> ok
[16:46] <xranby> is everything noted down in the pad
[16:46] <xranby> cheers
[16:46] <xranby> thank you for attending
[16:47] <xranby> looking forward to a rapid openjdk adoption
[16:47] <xranby> of new upstream openjdk releases
[16:47] <xranby> thank you as well at #uds for a nice session
[16:47] <xranby> and thank you gnu_andrew for giving the Fedora Red Hat input on things!
[16:49] <udsbotu> uds-gb-f: 5 minutes left in this session!
[16:50] <gnu_andrew> xranby, I wouldn't claim to represent Fedora.  I barely even use it ;-)  More upstream.
[16:50] <udsbotu> uds-gb-f: 4 minutes left in this session!
[16:50] <xranby> gnu_andrew: ok, thank you for correcting me..
[16:51] <gnu_andrew> xranby, just don't want the responsibility as speaking for Fedora.  Nowhere near qualified :-)
[16:51] <udsbotu> uds-gb-f: 3 minutes left in this session!
[16:52] <xranby> last reminder: Use JamVM if you run into any issues with hotspot
[16:52] <udsbotu> uds-gb-f: 2 minutes left in this session!
[16:53] <udsbotu> uds-gb-f: 1 minute left in this session!
[16:54] <gnu_andrew> xranby, not cacao?
[16:54] <xranby> and cacao
[16:54] <xranby> of course
[16:54] <udsbotu> uds-gb-f: This session has ended.
[16:55] <xranby> if you got a Raspberry PI use the cacao JIT: https://docs.google.com/a/gudinna.com/spreadsheet/oimg?key=0ArR4FJuzvXEsdDZkek5QcXd3c2ROMm1IajE1WmFwYkE&oid=3&zx=ipo09v1y3q53
[16:55] <gnu_andrew> xranby, does http://summit.ubuntu.com/uds-q/meeting/20495/foundations-q-java7/ work for you?  Logged in via my account but just taking me round in circles
[16:56] <xranby> gnu_andrew: yes it work for me
[16:56] <xranby> i can paste you a summary
[16:57] <xranby> gnu_andrew: http://fpaste.org/jkjV/
[17:02] <SpamapS>  ______  _____ __   __            _____  _______
[17:03] <SpamapS> |  ____||_   _|\ \ / /     /\    |  __ \|__   __|
[17:03] <SpamapS> | |__     | |   \ V /     /  \   | |__) |  | |
[17:03] <SpamapS> |  __|    | |    > <     / /\ \  |  ___/   | |
[17:03] <SpamapS> | |      _| |_  / . \   / ____ \ | |       | |
[17:03] <SpamapS> |_|     |_____|/_/ \_\ /_/    \_\|_|       |_|
[17:03] <SpamapS>  _  __ _______  _    _ __   __ ____            _____
[17:03] <SpamapS> | |/ /|__   __|| |  | |\ \ / /|  _ \    /\    |_   _|
[17:03] <SpamapS> | ' /    | |   | |__| | \ V / | |_) |  /  \     | |
[17:03] <SpamapS> |  <     | |   |  __  |  > <  |  _ <  / /\ \    | |
[17:03] <SpamapS> | . \    | |   | |  | | / . \ | |_) |/ ____ \  _| |_
[17:03] <SpamapS> |_|\_\   |_|   |_|  |_|/_/ \_\|____//_/    \_\|_____|
[17:04] <SpamapS> Who is there? Nobody?
[17:05] <SpamapS> Perfect! ^^
[17:05] <doko> you are watched ;p
[17:06] <ansgar> doko: But there is no video stream. How can you watch? :)
[17:07] <SpamapS> We need Ben Howard and Robie Basak as well, I don't hear them
[17:07] <smoser> hello?
[17:08] <DonKult_uds> smoser: hello, where are you?
[17:08] <smoser> i left session. i ddin't actually know this got scheduled.
[17:08] <smoser> if you want to discuss, we can, i can come
[17:09] <smoser> but i have no new information since wednesday morning.
[17:09] <SpamapS> bring utlemming and rbasak
[17:09] <smoser> k. i'll come down there.
[17:09] <DonKult_uds> i have… i didn't realized that another session was sheduled so started a discussion on dak -> https://lists.debian.org/debian-dak/2012/05/msg00006.html
[17:11] <SpamapS> can perhaps be moved to the next hour.. there are not many sessions for 11:00 - 11:55
[17:16] <SpamapS>        apt | 0.8.16~exp12ubuntu10 |       precise | source, amd64, armel, armhf, i386, powerpc
[17:20] <SpamapS> not apt pinning
[17:21] <SpamapS> host pinning
[17:27] <SpamapS> We're giving mirror admins the *option* to run the hash generation if they update aggressively.
[17:29] <ansgar> Requiring all mirrors to update seems not a good idea.  That might take quite a while.
[17:30]  * mvo is now listening
[17:32] <SpamapS> You can be atomic with 1 symlink mv atomicity and using hardlinks but nobody does that.
[17:33] <SpamapS> Better to have the format intrinsically made to handle skew between fetch of different files.
[17:33] <ansgar> You could also poulate a dists-new/ and then replace the old one.
[17:33] <SpamapS> ansgar: right, thats the hardlink method (since storing two mirrors would be a bit wasteful)
[17:34] <ansgar> Though it wouldn't help with clients that have the old Release and then get the new Packages. Which I believe what you want.
[17:34] <mvo> I think if we can we should do both, right? improve the server side mirror script plus add code to apt to make it more robust
[17:34] <mvo> if its orthogonal we should do both, right?
[17:34] <ansgar> (I never get those hash mismatches btw. ;) )
[17:36] <SpamapS> ansgar: We each do 100 apt-get updates a day during the dev cycle and we see them a lot.
[17:37] <SpamapS> Right, multiarch added more files to fetch
[17:38] <smoser> to re-iterate. we are interested in a solution.
[17:38] <SpamapS> BTW, somewhat related data point.. work is beginning this cycle of Ubuntu to combine main and universe into just 'main' (support will be communicated via seeds, not component)
[17:39] <mvo> sorry if I missed stuff, but dosn't option B with the timestamp/version in the http path fixes the proxy issue not too?
[17:39] <mvo> i.e. it will be different for each Packages file, no?
[17:39] <mvo> or am I missing something here?
[17:39] <SpamapS> We helped w/ the squid issue in Ubuntu btw by setting Expires to teh same value for Release and Packages*
[17:39] <udsbotu> uds-gb-f: 5 minutes left in this session!
[17:39] <smoser> mvo, it would fix the proxy issue, yes.
[17:40] <mvo> thanks smoser, so what is the problem with option (b) then? I'm not fully sure that I understand the current discussion? note that I'm not favoring one particular one, I just try to understand the issue
[17:40] <udsbotu> uds-gb-f: 4 minutes left in this session!
[17:41] <mvo> and also why do we need to change ever client? it would be libapt doing it, no?
[17:41] <ansgar> mvo: debmirror and friends might need changes as well I assume. At least if they want to take advantage of it.
[17:41] <udsbotu> uds-gb-f: 3 minutes left in this session!
[17:41] <mvo> thanks, I think I understand
[17:42] <SpamapS> Do we have actions from this btw? Or are the actions from the main blueprint enough?
[17:42] <mvo> ok
[17:42] <mvo> you could have a smart proxy that would ignore de.archive.ubuntu.com, archive.ubuntu.com etc as the hashsum is the only relevant piece
[17:42] <mvo> that is pretty clever
[17:42] <udsbotu> uds-gb-f: 2 minutes left in this session!
[17:43] <mvo> ansgar: right, the mirror stuff is client work too
[17:43] <udsbotu> uds-gb-f: 1 minute left in this session!
[17:44] <udsbotu> uds-gb-f: This session has ended.
[17:44] <SpamapS> ok, actions: move discussion forward in mailing list, and develop PoC
[17:45] <smoser> ansgar, you're right . debmirror would need improvements to know about either format acutally.
[17:45] <smoser> but with by-hash, i can update my mirror without the upstrem mirror to me implementing it.
[17:45] <smoser> (and keep the signed release)
[17:45] <SpamapS> I added the work item for rbasak to http://pad.ubuntu.com/uds-q-servercloud-q-apt-improvements
[17:47] <mvo> proof of concept for server? client? both?
[17:47] <mvo> apt-get / apt-ftparchive?
[17:48] <smoser> mvo both.
[17:48] <mvo> cool
[17:48] <SpamapS> I would think it has to be on both
[17:48] <smoser> server is almost implemented in the ehterpad
[17:48] <smoser> (ie, its really easy!)
[17:49] <smoser> ansgar, mvo so, just so everyone here is clear, myself, and rbasak are most interestd in getting a solution.
[17:49]  * mvo looks
[17:49] <mvo> right, we are all on the same page I think :)
[17:49] <mvo> we want to fix it too
[17:49] <smoser> http://paste.ubuntu.com/976863/
[17:50] <mvo> I think at least the apt-get part is pretty simple too for by-hash (well, for both)
[17:51] <mvo> a flag would be nice in the release file in both methods
[17:51] <mvo> but I guess that is hard for the mirrors that have not converted
[17:52] <ansgar> mvo: You need to check by downloading in any case.  You might have the new Release, but the mirror script might not mirror the new files.
[17:52]  * mvo nods
[17:54] <ansgar> Installers are also mentioned in Release. (A file with checksums actually.)
[17:54] <ansgar> Would the schemes store those as well in by-hash?
[17:55]  * mvo gtg
[17:55] <SpamapS> its only the lists
[17:56] <smoser> ansgar, i'm sorry, i didn't follow your question, i think/hope SpamapS did.
[17:57] <SpamapS> by-hash/2029447895 does not instill as much confidence as by-hash/affaf8faf5a5f8a9cc9c99e9b9b9ffa70900aff76f6f
[17:58] <SpamapS> crc32
[17:58] <SpamapS> ftw
[17:58] <SpamapS> ;)
[17:59] <SpamapS>  ____________________________________
[17:59] <SpamapS> / enough comedy, time to go to other \
[17:59] <SpamapS> \ sessions! Cheers                   /
[17:59] <SpamapS>  ------------------------------------
[17:59] <SpamapS>         \   ^__^
[17:59] <SpamapS>          \  (oo)\_______
[17:59] <SpamapS>             (__)\       )\/\
[17:59] <SpamapS>                 ||----w |
[17:59] <SpamapS>                 ||     ||
[18:02] <DonKult_uds> ansgar: thanks for coming on such a short notice! Any preference yet? ;)
[18:03] <smoser> https://wiki.ubuntu.com/QuantalQuetzal/ReleaseSchedule is our release schdule
[18:03] <smoser> ansgar, thank you.
[18:05] <ansgar> DonKult_uds: Hmm, not yet.  I'm not sure what would be better to use for the mirrors (and I don't have much experience with the mirror part).
[18:49] <udsbotu> uds-gb-f: 5 minutes left in this session!
[18:50] <udsbotu> uds-gb-f: 4 minutes left in this session!
[18:51] <udsbotu> uds-gb-f: 3 minutes left in this session!
[18:52] <udsbotu> uds-gb-f: 2 minutes left in this session!
[18:53] <udsbotu> uds-gb-f: 1 minute left in this session!
[18:54] <udsbotu> uds-gb-f: This session has ended.
[19:05] <yaili_> hi
[19:06] <yaili_> hi david
[19:06] <yaili_> no worries :)
[19:11] <dpm> http://summit.ubuntu.com/uds-q/meeting/20729/community-p-appdevelopers-site-incremental-improvements/
[19:42] <yaili_> I'd be happy to give feedback
[19:42] <yaili_> yes
[19:42] <yaili_> yeah, go ahead
[19:42] <yaili_> "Consultancy" :D
[19:50] <yaili_> makes sense, I think the idea in the first phase was to just get people going, so the focus on get started was more important
[19:50] <yaili_> now that people might know the site, makes sense to make make news more prominent
[19:50] <yaili_> I think it will need a bit of UX
[19:51] <yaili_> I will add a note to the web team that would be good for a UX to look at it
[19:52] <yaili_> ok, I'll talk to them on Monday, I'll present both options: a) we do it, b) someone helps us and we're feedback/help
[19:52] <yaili_> sure
[19:55] <udsbotu> uds-gb-f: 5 minutes left in this session!
[19:56] <udsbotu> uds-gb-f: 4 minutes left in this session!
[19:57] <udsbotu> uds-gb-f: 3 minutes left in this session!
[19:58] <udsbotu> uds-gb-f: 2 minutes left in this session!
[19:59] <udsbotu> uds-gb-f: 1 minute left in this session!
[19:59] <yaili_> thanks o/
[20:00] <udsbotu> uds-gb-f: This session has ended.
[22:02] <SpamapS>            _       __  __   ____    _____  _______
[22:02] <SpamapS>     /\    | |     |  \/  | / __ \  / ____||__   __|
[22:03] <SpamapS>    /  \   | |     | \  / || |  | || (___     | |
[22:03] <SpamapS>   / /\ \  | |     | |\/| || |  | | \___ \    | |
[22:03] <SpamapS>  / ____ \ | |____ | |  | || |__| | ____) |   | |
[22:03] <SpamapS> /_/    \_\|______||_|  |_| \____/ |_____/    |_|
[22:03] <SpamapS>  _____    ____   _   _  ______
[22:03] <SpamapS> |  __ \  / __ \ | \ | ||  ____|
[22:03] <SpamapS> | |  | || |  | ||  \| || |__
[22:03] <SpamapS> | |  | || |  | || . ` ||  __|
[22:03] <SpamapS> | |__| || |__| || |\  || |____
[22:03] <SpamapS> |_____/  \____/ |_| \_||______|
[22:06] <SpamapS> This is my dream session.. Don't take it away from me! Upstart and Juju in perfect harmony... ;)
[22:06] <SpamapS> I printed out materials you can find them under your chairs...
[22:07] <jamespage> ?
[22:07] <SpamapS> Yes sort of
[22:07] <SpamapS> I did not file this one
[22:08] <SpamapS> The idea is to make sure that its really easy to use an upstart job in your charm
[22:08] <SpamapS> Yeah, and I think we skipped this topic
[22:08] <SpamapS> Its not really a 1 hour conversation...
[22:09] <SpamapS> Just a "anybody have ideas on how to make it easy and/or encourage it"
[22:09] <SpamapS> Upstart jobs are so simple..
[22:10] <SpamapS> IMO upstart job is the obvious way to start a service deployed by a charm.
[22:10] <SpamapS> Nor am I. :)
[22:11] <SpamapS> Here's what I think.. we talked about "flag bearing charms" that implement all best practices. We can pick 1 or 2 that do it right, and make sure they're known as such.
[22:11] <SpamapS> +1's all around?
[22:12] <jamespage> +1
[22:12] <SpamapS> The start/stop hooks are horribly misunderstood
[22:12] <SpamapS> I think you just got it wrong ;)
[22:13] <SpamapS> Even w/ an upstart job, you need to have the action required to start/stop it in start/stop hooks
[22:13] <SpamapS> start/stop can be called w/o rebooting the machine
[22:14] <SpamapS> No templates needed.. most jobs are 3 or 4 lines
[22:14] <SpamapS> james hunt.. I have a question on a potential feature
[22:15] <jamespage> so good examples and some extra information as comments in the hooks to help people understand what does what
[22:15] <SpamapS> lxc containers cannot be run without network abstraction because of upstart's private netlink socket ...
[22:16] <SpamapS> can we create some way for upstart to detect that it is running in a container that does not have network namespaced, and listen somewhere else?
[22:16] <SpamapS> (this is unrelated to the blueprint topic)
[22:17] <SpamapS> no, /sbin/init fails
[22:18] <SpamapS> yes but only if your container has the network namespaced
[22:18] <SpamapS> if you let the container inherit the host's networking, /sbin/init cannot listen on the netlink socket.
[22:18] <SpamapS> AH
[22:19] <SpamapS> ok, because we wanted to use LXC to contain juju units without namespacing the network, but we can't
[22:20] <SpamapS> Ok, thought it was upstart's problem. Ok, thats all from me. :)
[22:20]  * SpamapS un-hijacks the session
[22:21] <SpamapS> For ~charmers .. grand-ballroom-a would be a good spot to go to see how ubuntu is unblocking their review/contributor processes
[22:21]  * SpamapS goes there now
[22:22] <SpamapS>   ____   _   _  ______   __  __   ____   _____   ______
[22:22] <SpamapS>  / __ \ | \ | ||  ____| |  \/  | / __ \ |  __ \ |  ____|
[22:22] <SpamapS> | |  | ||  \| || |__    | \  / || |  | || |__) || |__
[22:22] <SpamapS> | |  | || . ` ||  __|   | |\/| || |  | ||  _  / |  __|
[22:22] <SpamapS> | |__| || |\  || |____  | |  | || |__| || | \ \ | |____
[22:22] <SpamapS>  \____/ |_| \_||______| |_|  |_| \____/ |_|  \_\|______|
[22:22] <SpamapS>   _____  ______   _____  _____  _____  ____   _   _
[22:22] <SpamapS>  / ____||  ____| / ____|/ ____||_   _|/ __ \ | \ | |
[22:22] <SpamapS> | (___  | |__   | (___ | (___    | | | |  | ||  \| |
[22:22] <SpamapS>  \___ \ |  __|   \___ \ \___ \   | | | |  | || . ` |
[22:22] <SpamapS>  ____) || |____  ____) |____) | _| |_| |__| || |\  |
[22:22] <SpamapS> |_____/ |______||_____/|_____/ |_____|\____/ |_| \_|
[22:54] <udsbotu> uds-gb-f: 5 minutes left in this session!
[22:55] <udsbotu> uds-gb-f: 4 minutes left in this session!
[22:56] <udsbotu> uds-gb-f: 3 minutes left in this session!
[22:57] <udsbotu> uds-gb-f: 2 minutes left in this session!
[22:58] <udsbotu> uds-gb-f: 1 minute left in this session!
[22:59] <udsbotu> uds-gb-f: This session has ended.
[23:20] <SpamapS> o/
[23:24] <xranby1> i am having issues connecting to the icecast audio server
[23:32] <SpamapS> sqlite does table locking, so multi-user will just be single threaded
[23:35] <SpamapS> sounds good :)
[23:43] <SpamapS> 1 year
[23:43] <SpamapS> 8.04 == 2008 + 5 == 2013
[23:54] <udsbotu> uds-gb-f: 5 minutes left in this session!
[23:55] <udsbotu> uds-gb-f: 4 minutes left in this session!
[23:56] <udsbotu> uds-gb-f: 3 minutes left in this session!
[23:57] <udsbotu> uds-gb-f: 2 minutes left in this session!
[23:58] <udsbotu> uds-gb-f: 1 minute left in this session!
[23:59] <udsbotu> uds-gb-f: This session has ended.
[23:59] <SpamapS> __          __ ______   __  __            _____   ______   _____  _______  _
[23:59] <SpamapS> \ \        / /|  ____| |  \/  |    /\    |  __ \ |  ____| |_   _||__   __|| |
[23:59] <SpamapS>  \ \  /\  / / | |__    | \  / |   /  \   | |  | || |__      | |     | |   | |
[23:59] <SpamapS>   \ \/  \/ /  |  __|   | |\/| |  / /\ \  | |  | ||  __|     | |     | |   | |
[23:59] <SpamapS>    \  /\  /   | |____  | |  | | / ____ \ | |__| || |____   _| |_    | |   |_|
[23:59] <SpamapS>     \/  \/    |______| |_|  |_|/_/    \_\|_____/ |______| |_____|   |_|   (_)