xnox | i've synced db6.0 twice now. so the older one can be dropped. but i'd like the ^ one to be accepted please =) | 08:12 |
---|---|---|
cjwatson | xnox: done | 08:13 |
jamespage | Daviey, how would the SRU team feel about backporting the 1.9.0 openvswitch from raring to 12.04 to support the lts-raring enablement kernel? | 08:13 |
xnox | cjwatson: cheers! | 08:14 |
jamespage | it supports 3.8 natively and I feel its actually much lower risk and trying to backport the multitude of cherry-picks that don't apply cleanly to the 1.4.x codebase we have in 12.04 | 08:14 |
Daviey | jamespage: Hmm, a dedicated NEW versioned package name (like linux), or bumping the version? | 08:18 |
jamespage | Daviey, hmm | 08:21 |
jamespage | Daviey, I was thinking a straight backport to updates | 08:23 |
Daviey | jamespage: Dual supporting both is more favourable IMO, but lack of dynamic dependencies makes this kinda hellish.. I can't think of a good user experience fix for handling this gracefully | 08:23 |
jamespage | Daviey, the other option which is not ideal is to provide 1.9.0 via precise-backports | 08:23 |
jamespage | so that 1.4.0 continues to exist for those that want to stick with 3.2 | 08:23 |
jamespage | but lts-raring users have to use the backports version.... | 08:24 |
jamespage | I'd need to run that past the backports team obviously | 08:24 |
Daviey | jamespage: Probably a good idea to bring this to the TB mailing list, outlining the issue and possible solutions | 08:25 |
jamespage | actually that might work better - I suspect that for the saucy enablement kernel I'll need to backport 1.10 which drops the brcompat module | 08:25 |
jamespage | Daviey, OK - I'll send something this morning | 08:25 |
Daviey | jamespage: Well, straight backports cannot really solve this, can it? Unless you still do the compat work in the current openvswitch package? | 08:26 |
jamespage | Daviey, ? | 08:26 |
jamespage | its already backwards compatible | 08:26 |
jamespage | i.e. 1.10 works on the 3.2 kernel | 08:27 |
Daviey | jamespage: Current ovs in precise works with current lts kernel, but not raring backport? | 08:27 |
Daviey | Oh! | 08:27 |
jamespage | Daviey, yes | 08:27 |
jamespage | The version we have in precise is good up to the quantal hwe kernel | 08:27 |
jamespage | but that was painful and we have to carry a large number of patches | 08:27 |
jamespage | (and I got it wrong first go) | 08:28 |
Daviey | jamespage: Right.. and you want to avoid that again... and I agree the regression potential of trying to cherry pick that much is a problem. | 08:29 |
Daviey | So what i am saying is, making use of backports will mean that -updates doesn't work with the most recent HWE kernel. | 08:29 |
Daviey | jamespage: AND, we ned to move quite quickly... 12.04.3, scheduled for next Thursday (is that still correct?).. will not work with openvswitch as it has the raring kernel, right? | 08:31 |
jamespage | Daviey, no | 08:31 |
Daviey | I'm wrong? | 08:32 |
Daviey | jamespage: Am i wrong? | 08:40 |
jamespage | Daviey, the version of openvswitch in 12.04 will not work with the raring HWE kernel | 08:41 |
jamespage | well the user space tools might with the native kernel module | 08:41 |
jamespage | but the openvswitch-datapath-dkms module which is the one everyting uses will not | 08:41 |
Daviey | jamespage: Right.. the native support doesn't have GRE support, right? | 08:42 |
jamespage | Daviey, thats correct | 08:42 |
jamespage | Daviey, hmm - I wonder whether we can skew the userspace tools against the dkms module version | 08:43 |
jamespage | Daviey, that way we could provide a versioned openvswitch-datapath-dkms-lts-raring package only | 08:43 |
Daviey | Yeah, I think that is what I was thinking earlier.. but create a good user experience as we cannot do dynamic depends | 08:44 |
Daviey | err, creates a poor* | 08:44 |
jamespage | Daviey, well we have todo an explicit install of the dkms package anyway | 08:52 |
jamespage | Daviey, so its not that much different | 08:52 |
Daviey | jamespage: true.. Yeah, I think that is the best balance | 08:54 |
jamespage | Daviey, just testing that now | 08:55 |
Daviey | We get raring-hwe compatibility, with no risk of current users. | 08:55 |
Daviey | I'd also say that it is a manageable workflow in the future | 08:55 |
Mirv | could an archive admin add "poppler-qml-plugin" to the whitelist of source packages coming from cu2d? maybe cjwatson? both didrocks and seb128 are away today. | 08:57 |
Daviey | Mirv: Sorry, I would help.. but I do not know how cu2d works, and I have been unable to find any docs. | 09:08 |
jamespage | Daviey, userspace tools for 12.04 appear to be functional with newer dkms versions! | 09:08 |
Mirv | Daviey: yeah the slight problem here is that I don't know where/what the whitelist didrocks touches is, but cyphermox told it's on the archive machines and not part of cu2d system (https://wiki.ubuntu.com/DailyRelease) as such | 09:10 |
Daviey | jamespage: Super, that is good news. I'd be happy processing this approach myself, unless anyone else objects. | 09:10 |
jamespage | Daviey, great - I'll not bother emailing the TB as we have a low risk approach now | 09:11 |
jamespage | marvellous | 09:11 |
* jamespage thinks its good to talk these things through! | 09:11 | |
Daviey | Mirv: Yeah, I'm not even sure I can do that.. I'd rather not poke around, and risk getting it wrong. Sorry. :( | 09:13 |
Mirv | Daviey: I totally understand, that's why the idea of getting that Thing documented is a good one :) | 09:15 |
Laney | It's something that gets committed to bzr and then pulled on lillypilly | 09:20 |
* Laney waves hands and flails | 09:20 | |
sil2100 | Daviey: hi! Maybe you could help us in some other way! Do you have a minute? :) | 09:43 |
sil2100 | Daviey: in the daily-release process, we have a requirement that whenever there is a packaging change in a package that is ready for releasing, we need and archive-admin-or-similar to ACK the change - so that no 'risky' packaging change gets released into Ubuntu | 09:44 |
sil2100 | Daviey: normally didrocks or seb128 do those, but now they're both gone | 09:44 |
sil2100 | Daviey: we have one thing like this now, it's really a one-liner that we already acknowledged, but we still need some core-dev to ACK it as well | 09:46 |
sil2100 | Daviey: I know it's hard to ACK something when you don't know the code, but it's just to make sure nothing unwanted gets into the archive | 09:47 |
sil2100 | Daviey: https://jenkins.qa.ubuntu.com/view/cu2d/view/Head/view/Indicators/job/cu2d-indicators-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_libusermetrics_1.1.1+13.10.20130816-0ubuntu1.diff | 09:47 |
sil2100 | Could you take a look? ;) | 09:47 |
=== tkamppeter_ is now known as tkamppeter | ||
jamespage | Daviey, I've emailed upstream just to validate the approach we discussed | 09:55 |
jamespage | I might be missing something re userspace/kernel module compatibility | 09:55 |
Daviey | sil2100: Well, reviewing the packaging diff represented there - it looks like something i'd be happy sponsoring. | 09:56 |
sil2100 | Daviey: ! | 09:57 |
sil2100 | Daviey: thank you! | 09:57 |
sil2100 | Mirv: puublish! | 09:57 |
Daviey | jamespage: well.. this base approach gives us some known compatibility.. in the worsed case, we'd need to also add a raring-hwe user space tools - right? | 09:57 |
* sil2100 goes into slow-motion | 09:57 | |
jamespage | Daviey, yes | 09:57 |
sil2100 | Mirv: noooooow~! | 09:58 |
sil2100 | Daviey: thanks! | 09:59 |
Mirv | sil2100: oh, you were shouting here :) | 09:59 |
sil2100 | ;) | 10:00 |
cjwatson | Mirv: I'm at DebConf and I don't know much about cu2d anyway, sorry. | 10:19 |
cjwatson | sil2100: This process should involve verification by somebody with upload privileges, not by an archive admin (a much smaller and much more contended set of individuals) | 10:20 |
cjwatson | For the record | 10:20 |
Daviey | cjwatson: Yeah, i almost said that - but i think sil2100 knew that, as he quantified with "need some core-dev to ACK it" .. I assume it's not in a pkg set. | 10:31 |
Mirv | cjwatson: ok thanks anyway, and happy debconfing. and yes, the thing sil2100 asked for requires just core-dev. | 10:59 |
Mirv | it's good that didrocks is away for a small while before real vacation, so that we notice these beforehand :) | 10:59 |
sil2100 | Indeed :) | 11:15 |
sil2100 | Actually, once again we need a core-dev to ACK another packaging change | 11:15 |
sil2100 | A quick one again | 11:15 |
Laney | core-devs> #ubuntu-devel | 11:39 |
sil2100 | ogra_: hi! Are you still around? | 13:38 |
sil2100 | ogra_: we have another near-deadline packaging ACK that needs ACKing, maybe you could help this time as well ;) ? | 13:38 |
sil2100 | ogra_: https://jenkins.qa.ubuntu.com/view/cu2d/view/Head/view/Unity/job/cu2d-unity-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_libunity_7.0.11+13.10.20130816.2-0ubuntu1.diff | 13:39 |
ogra_ | sil2100, i assume this has been tested (cant really judge the sed lines so i need to truset a testrun you did ) | 13:41 |
ogra_ | *trust | 13:41 |
sil2100 | ogra_: it passed our testing (failures in the acceptance threshold) | 13:42 |
ogra_ | sil2100, fine, then, go ahead | 13:42 |
sil2100 | ogra_: thanks! You're a life saver | 13:42 |
jamespage | Daviey, ^^ updated iscsitarget for lts raring hwe kernel | 13:45 |
Daviey | jamespage: reviewed and ^ | 13:49 |
jamespage | Daviey, thanks muchly! | 13:49 |
=== rtg_ is now known as Guest11042 | ||
tkamppeter | jdstrand, hi | 14:52 |
jdstrand | tkamppeter: hey | 15:00 |
cyphermox | sil2100: Mirv: when you need a core-dev to review things before upload, that I or kenvandine can do | 15:22 |
cyphermox | so, for the cu2d magic that apparently is needed on lillypilly | 15:22 |
sil2100 | \o/ | 15:22 |
cyphermox | seems like that would be just a matter of a bzr pull for the cupstream2distro-config branch somewhere on the machine | 15:23 |
sil2100 | cyphermox: first of all, do you have a moment for a quick pkg ACK? See #ubuntu-desktop ;) | 15:23 |
sil2100 | cyphermox: awesome | 15:23 |
cyphermox | cron should be able to say exactly where, but I don't know how didrocks did it, and I don't have that kind of god powers | 15:23 |
Laney | needs documenting for bus factor reasons | 15:23 |
sil2100 | Ok | 15:23 |
sil2100 | Laney, cyphermox: https://wiki.ubuntu.com/DailyRelease/FAQ#Adding.2BAC8-removing_components_to_a_stack this documents some parts probably | 15:24 |
cyphermox | right | 15:25 |
cyphermox | it is documented, just in a non-obvious way | 15:25 |
cyphermox | I totally read over that without noticing >.< | 15:26 |
sil2100 | hehe | 15:35 |
sil2100 | Right, it's not being explicitly mentioned anywhere | 15:35 |
tkamppeter | jdstrand, did you have a look into the openjpeg MIR, bug 711061? | 16:10 |
ubot2` | Launchpad bug 711061 in openjpeg (Ubuntu) "[MIR] libopenjpeg2" [High,Confirmed] https://launchpad.net/bugs/711061 | 16:10 |
cyphermox | Laney: so, is that wiki page enough information? | 16:12 |
Laney | dunno | 16:12 |
Laney | I don't even know what is being requested. :P | 16:12 |
jdstrand | tkamppeter: no. let me see if I can get someone else to do it | 16:13 |
Laney | cyphermox: that branch is on 608 but trunk is 655 | 16:17 |
tkamppeter | jdstrand, OK. | 16:39 |
cyphermox | Laney: if you could bzr pull on that branch I suspect it will be enough | 17:07 |
Laney | I don't know what it means and so I'm reluctant to do so | 17:07 |
=== bjf is now known as bjf[afk] | ||
Sarvatt | anyone know whats up with x11-utils migration? autopkgtest for gvfs 1.17.2-0ubuntu6: RUNNING but it looks like it succeeded in jenkins unless i'm misunderstanding it | 21:00 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!