[02:11] <infinity> There.  Mosh works on ppc64el.  Port done.  Nothing else matters.
[14:16] <teward> has an official date been decided for the quantal EOL yet?
[20:54] <stgraber> arkose has been badly broken for about a year due to LXC changes, changes to su, kernel changes and the introduction of userns
[20:54] <stgraber> I won't have time to fix it this cycle, so I plan on getting it out of 14.04 entirely
[20:55] <infinity> :(
[20:55] <infinity> Poor arkose.
[20:55] <stgraber> the upload above does just that for Edubuntu (only seeder of those packages), I'm also going to push a matching installer slideshow update and then file and archive removal bug
[20:55] <stgraber> all the features I ever needed for arkose now exist but it needs pretty much a full rewrite to use them and that won't happen by release
[20:55] <stgraber> I still hope for a 2.x release which will use userns and unprivileged overlayfs to provide the exact same features but without requiring any privileges
[20:57] <infinity> stgraber: Maybe you can rewrite it in a language that doesn't suck next time. ;)
[20:59]  * infinity blinks at ubiquity-slideshow-ubuntu being in the kubuntu packageset.
[21:00] <stgraber> infinity: well, I don't feel like doing it in C, so my other options for the LXC binding is Go, Lua, Ruby or Python3. I think I'd stick to Python3 :)
[21:01] <infinity> stgraber: No perl?
[21:01] <infinity> (But yeah, C was what I was suggesting)
[21:01] <stgraber> infinity: bug 1299904 (I probably shouldn't be processing that one myself)
[21:01] <ubot2> Launchpad bug 1299904 in arkose (Ubuntu) "Please remove arkose from the archive for 14.04" [Undecided,New] https://launchpad.net/bugs/1299904
[21:01] <infinity> Seems weird to sandbox potentially tiny/fast C applications by running a massive python interpreter.
[21:02] <stgraber> I suspect 90% of arkose's code could now go away as it's since been re-implemented in liblxc itself so maybe doing it in C wouldn't be as horrific as it'd have been for the initial implementation :)
[21:02] <infinity> shell would work too, I suppose, if there were enough CLI utilities to make that work.
[21:02] <stgraber> the GUI and nautilus integrations would probably remain python, because those are definitely a pain to do in C :)
[21:03] <stgraber> yeah, shell would be an option, I "think" I could actually do that with the current LXC tools, definitely a fun thing to try if it wasn't for ENOTIME...
[21:04] <infinity> PS: Thanks for hiding your slideshow change in the middle of a translation update.
[21:04] <stgraber> infinity: sorry :(
[21:04] <infinity> Oh, hah, you just commented out the slide.  Kay.
[21:04] <stgraber> yep, that's the usual way. Removing it would also remove the translations which may one day be useful again.
[21:06] <infinity> stgraber: Yeah, seems sensible to me.