| micahg | teward: hi, I wanted to ask you about znc (or maybe you answered already) | 01:55 |
|---|---|---|
| teward | micahg: hm? | 01:55 |
| teward | probably haven't, i've been focusing on the nginx issue on the -server mailing list, which prompted a release team ping and has a pending TB ping when robie gets to it with his busy schedule | 01:56 |
| teward | what's up? | 01:56 |
| micahg | hi, so, there was one last thing WRT znc-dev, it's not checked, did you actually verify it according to what you mentioned as the test for it? | 01:56 |
| micahg | BTW, if you're actively supporting nginx, I'd be willing to put it in backports | 01:58 |
| teward | i've got another question for you in response to that, but first, znc. | 01:58 |
| micahg | ok :) | 01:58 |
| teward | micahg: to be completely honest, I haven't had time recently to poke it, I've had it on my radar for some time, but what ended up happening in the past couple months was work, database explosions, iSCSI SAN explosions... | 01:58 |
| teward | and me getting ill | 01:58 |
| teward | so ZNC's been on the list, but not in the priority list | 01:59 |
| micahg | sorry, are you at least better? | 01:59 |
| teward | yeah, i am | 01:59 |
| micahg | that's good to hear | 01:59 |
| teward | but then the iSCSI at work died | 01:59 |
| micahg | sigh | 01:59 |
| teward | so abf;wehrfaweruhamat *implode* | 01:59 |
| teward | anyways, nginx. | 01:59 |
| micahg | ok, so I'll assign znc to you then so it's not on my list :) | 01:59 |
| teward | micahg: tricky tricky part is naxsi was dropped | 01:59 |
| micahg | whenever you get to it is fine | 01:59 |
| teward | micahg: ack on ZNC being assigned back to me | 01:59 |
| teward | micahg: so with regards to nginx, unless you're OK with nuking nginx-naxsi as part of backports, there's a limit to how far we can backport. And from where | 02:00 |
| micahg | if it's not part of the new upstream, I think that's fine | 02:00 |
| micahg | nothing depends on it in trusty at least | 02:00 |
| teward | eheheheh... | 02:01 |
| micahg | (directly) | 02:01 |
| teward | um... | 02:01 |
| teward | nginx-naxsi-* might error | 02:01 |
| micahg | oh, is it a separate source? | 02:01 |
| teward | no, it's included in the same tarball | 02:01 |
| teward | flashback to Trusty LTS time, and the MIR for nginx, we created nginx-core for main | 02:01 |
| micahg | nothing seems to depend on that either | 02:02 |
| teward | but nginx-extras, nginx-full, etc. are still present in the source package | 02:02 |
| teward | and nginx-naxsi* was in at until pre-15.04 merge/sync | 02:02 |
| teward | when Debian dropped | 02:02 |
| micahg | trusty: http://paste.ubuntu.com/11839254/ | 02:02 |
| teward | micahg: that's just an rdeps check, or a specific script? | 02:02 |
| micahg | reverse-depends | 02:03 |
| teward | micahg: currently a backport wouldn't make sense except vivid -> trusty, to put 1.6.x in, but we're currently looking at 1.9.x for Wily and we may want to push nginx backport considerations until after Wily | 02:03 |
| * micahg had the AA script somewhere, just can't remember right now where | 02:03 | |
| teward | micahg: currently, the discussion jumps forward to LTS, to determine course-of-action for wily which will impact X | 02:03 |
| micahg | ok, that's fine, I just wanted to offer that if someone is actively maintaining it and willing to test, I don't mind keeping it in -backports | 02:04 |
| teward | and the current discussion / proposal is to merge 1.9.x for Wily, merge for X, and then post-release/post-freeze get 1.10.x in | 02:04 |
| teward | micahg: yeah, i'm poking it, but I use mostly Trusty | 02:04 |
| micahg | yeah, I read the list, plan sounds sound | 02:04 |
| teward | the packaging as exists in Debian exists in the PPAs which I still maintain, so I know it works | 02:04 |
| teward | but I tested on Trusty | 02:05 |
| teward | not Vivid | 02:05 |
| teward | so backporting nginx? Great idea! Is it on my radar right now? Not really. | 02:05 |
| teward | (because TB input is needed, the Wily decision for it needs to be made, and a few other things nginx-related for W and X series I need put in place) | 02:05 |
| micahg | right, that's the funny thing with backports, need to test the in between releases to preserve upgrades, I think when utopic is EOL, we can backport directly to trusty, I have to double check that upgrades aren't enabled from T -> V though | 02:06 |
| micahg | not sure if anything changed in the new 9 month release world | 02:06 |
| teward | i thought LTS->LTS is enabled, but not LTS->[skip]->nonLTS | 02:06 |
| micahg | right, that's what I think too | 02:06 |
| teward | at least, AIUI | 02:06 |
| teward | i could test it, tomorrow, at work, fake-run `do-release-upgrade` in sandbox mode | 02:07 |
| teward | 'test' the upgrade, but don't actually upgrade | 02:07 |
| teward | and see what it says the version is | 02:07 |
| teward | i *think* you still need the inbetweens | 02:07 |
| teward | but meh | 02:07 |
| micahg | well, utopic is supported for another 2.5 weeks :) | 02:07 |
| micahg | I can test it myself when utopic is EOL | 02:07 |
| teward | at which point .kill == true. | 02:07 |
| teward | :p | 02:07 |
| teward | i know someone on Vivid uses the packaging, but not me. | 02:08 |
| micahg | ok | 02:09 |
| micahg | I'd like to find a way to get some of the migration checks available to backports so we might be able to automate some of the testing, I've been thinking about that a bit | 02:10 |
| teward | mmm | 02:11 |
| micahg | proposed -> release migration checks | 02:11 |
| teward | well, my current task is, other than prepping the options for the TB for nginx in W and X, is to figure out whether we can get better bug reporting data from Apport for post-installation-failed bugs... since systemd made the available output from the 'start' of the service total crap | 02:12 |
| teward | so that has to be looked at before backports are even looked at on my radar | 02:12 |
| micahg | pitti would probably be a good person to talk to as he's worked with both | 02:12 |
| teward | indeed. | 02:12 |
| teward | i poked pitti earlier in -devel | 02:12 |
| teward | didn't get a return ping | 02:12 |
| teward | timezone differential perhaps | 02:13 |
| teward | and the discussion/minirant i had in -server led me to the same conclusion: talk to pitti | 02:13 |
| teward | the tricky part: get apport to execute two commands, and keep the output, and attach it. | 02:13 |
| micahg | he'll be online in 2-3 hours (assuming he's not on vacation) | 02:13 |
| teward | and i haven't found any such apport hooks. | 02:13 |
| teward | i'mma be asleep in 2-3 hours | 02:14 |
| teward | because 22:15 here now | 02:14 |
| micahg | heh | 02:14 |
| teward | and i need sleep :p | 02:14 |
| teward | just waiting to pump a little more power in this system since it ran to 2% battery thanks to a power outage due to storms. and the mirrors team poked me and almost lost me due to no power :P | 02:14 |
| teward | no, but i'll stop by early tomorrow morning my time, see if I can't find pitti, and at worst drop an email or such | 02:15 |
| teward | since this has been an issue since the 15.04 release for me | 02:15 |
| teward | not to mention all those VPS providers' images having Apache preinstalled adds to the hell | 02:16 |
| micahg | yeah, he's usually around till at least the US morning | 02:16 |
| micahg | teward: have you seen https://wiki.ubuntu.com/Apport#Per-package_Apport_Hooks | 02:17 |
| teward | micahg: the one thing I really don't want to hear is "Yeah, apport hooks can't do that..." in which case the solution is to (a) create additional binaries for nginx, basically $nginxflavor-manual that doesn't start the service on postinst | 02:17 |
| teward | but that quadruples build time and build space | 02:17 |
| teward | micahg: yeah, to no avail | 02:17 |
| teward | what i think i need is a custom one | 02:17 |
| teward | or better documentation on how to craft apport hooks | 02:18 |
| teward | because it's not logfiles we're including, we're going to need to capture some of the output from `systemctl status nginx.service`, `journalctl -xe`, and also grab the /var/log/nginx/error.log files if they exist (depending on where the failure happens) | 02:19 |
| * teward shrugs | 02:20 | |
| micahg | http://bazaar.launchpad.net/~apport-hackers/apport/trunk/view/head:/doc/package-hooks.txt, it's python, so sky should be the limit | 02:20 |
| teward | i've been beating this on and off the past two months with no success | 02:20 |
| teward | i need to brush up on python too xD | 02:20 |
| teward | ehh i'm too tired to actually work on it now | 02:22 |
| teward | micahg: no, but feel free to reassign the ZNC backport to me, i think the znc-dev libicu issue is still present | 02:25 |
| micahg | ok, I reassigned | 02:25 |
| teward | yeah, it's 'committed' but not fixed | 02:25 |
| teward | so we may want to either wait until that's fixed, since I hear a 1.6.1 bugfix release may be coming soon for ZNC | 02:25 |
| teward | or fix it ourselves | 02:26 |
| micahg | ok, well, I'll leave that to you, I can sponsor stuff at a later time for you if you want to fix it or wait (in which case, it might be easier to close the backport request and wait 2.5 weeks for utopic to be EOL) | 02:26 |
| micahg | I guess if it's a bug affecting vivid as well, that's SRUable | 02:27 |
| micahg | (at least a targeted fix for the bug) | 02:27 |
| teward | yeah that was my thought, but we may want to poke the release team on that first. | 02:27 |
| teward | because it adds a dep to the actual package itself | 02:28 |
| teward | fixes the one bug, but not sure if it's entirely SRUable | 02:28 |
| teward | but again, i need sleep xD | 02:28 |
| teward | can't do any major devel consideration stuff while tired | 02:28 |
| teward | last time i did that i ended up writing five pages of unintelligible python scripts | 02:29 |
| dholbach | good morning | 06:45 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!