roaksoaxhowdy! I have a suspicion that squid3 is broken in https://launchpad.net/ubuntu/+source/squid3/3.5.12-1ubuntu100:40
roaksoaxit is currently in proposed, but if it hits main it may break eveyrbody00:41
roaksoaxor maybe just a false alarm00:43
roaksoaxseems broken00:45
infinityroaksoax: You're not wrong.  See the blocking bug.  I was experimenting locally with fixing it.01:46
infinitystgraber: If by "looking at lxd" you mean "passing out and sleeping the afternoon away", then yes.  That's what I was doing.01:47
infinitystgraber: You misspelled CIDR as CDR.  Also, wow, that shell function is something else.05:24
stgraberinfinity: well at least I did it consistently, I'm calling it with the typo too :)05:28
infinitystgraber: Indeed. :)05:29
infinitystgraber: Oh, nice, you have an inverse of the function as well.  I need to save these for later use.05:29
infinityAnd the inverse is spelled correctly. :P05:29
stgraberyeah, we store cidr only in debconf (because it's 2016) but the script we use to bring the bridge up is stuck a decade ago so its config file needs it as a good old mask05:30
infinityWell, that was obviously unreviewable according the general laws of code review, but nothing obviously broken jumped out.05:30
stgraberand then since we want to be nice and load whatever the user wrote over the file we generated, we have to do the conversion the other way around when feeding debconf with whatever's in the file05:31
stgraberinfinity: yeah, took me about a week to get all that shit right, or at least sufficiently right that all my test cases passed :)05:32
stgrabernice, now we just need to get the Juju folks to merge tych0's branch to support that thing and then we can flip the switch and stop having lxd pull all of lxc and lxcbr0 everywhere. That should shrink the server ISO and cloud images a bit and most importantly, those two won't have a lxcbr0 bridge around that most people won't use (and that's potentially masking some useful subnet)!05:33
infinitystgraber: And yeah, "it's 2016" is usually my claimed excuse for preferring CIDR too, but the real reason is just that zero through thirty-two is a lot simpler than remembering the binary mask intervals.05:34
stgraberyeah, not too bad so long as you deal with just /8, /16, /24 and /32 but anything else is a bit painful :)05:36
infinityMost of my subnets are .224, .240, and .248, and counting backwards in binary hurts my brain.05:36
stgraberwell, I'm mostly dealing with IPv6 these days and I'm quite happy we have CIDR because as much as some network gear is trying to convince me that 2607:f2c0:f00f:2700:e88e:6217:1c70:5a55/ffff:ffff:ffff:ffff:0000:0000:0000:0000 is valid, it's really long and stupid05:38
infinityHahaha, yes.05:39
stgraber(got a new VDSL modem which doesn't do CIDR for IPv6 and also doesn't do shortening, so :: didn't save me either, I had to type every single zero...)05:39
Logando completely new packages need an FFe? or will they just be reviewed in the NEW queue?06:53
Loganalso I feel like dogecoin should not be in Xenial for the same reasons as the bitcoin removal07:28
=== zequence_ is now known as zequence
tewardLogan: I agree with your assessment, though I would expand that to include any cryptocurrency14:30
teward(though I have no real say, I believe)14:30
infinityLogan: Good catch.  Removed and blacklisted.18:44
tewardinfinity: we don't have any other cryptocurrencies lying around do we?18:47
infinityteward: Dunno.18:49
infinityteward: A quick search on "coin" also pulls up cgminer and bfgminer, but I don't know enough about bitcoin to know if "miners" are as useless as wallets/client in the face of network changes.18:49
infinityI certainly see no whiney Debian bugs about either one.18:50
tewardinfinity: i don't think the miners are an issue, they only rely on hashing algos18:50
tewardrather than the actual cryptocurrencies18:50
tewardthough i'm still partial to removing those because new mining hardware won't be caught by older miners18:51
tewardthough that logic would apply to Trusty being removed completely from existence if we wanted every package (kernel) to support every new hardware18:52
teward(this is what HWE stacks were for)18:52
infinityYeah, I don't think removing on those grounds would make sense.18:54
tewardat first glance into Unstable I don't see anything but bitcoin,litecoin,dogecoin18:55
tewardand if those are all removed and blacklisted, then we should be fine18:55
tewardand we leave the miners alone unless we find some big reason to remove18:55
infinityYeah, those are all blacklisted.18:55
tewardthen I think we're all set :)18:57
Loganinfinity: thanks!21:12

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!