falcojr | IIRC most warnings come from the dependencies rather than the code itself, but if there's some easy ones to remove, sure! It'd be best to do as a separate commit in a separate PR though | 00:15 |
---|---|---|
=== david__ is now known as davido | ||
Odd_Bloke | falcojr: Would you be able to take a look at https://github.com/canonical/cloud-init/pull/837 ? It sounds like we may have regressed vendor-data? | 14:51 |
falcojr | yep, I'll take a look | 15:01 |
falcojr | I commented on the PR, but I don't think this is a regression. Same behavior happens if I jump to (much) older releases of cloud-init | 16:04 |
Odd_Bloke | falcojr: Would any code set vendordata_raw to a list in those older versions? | 16:13 |
falcojr | I inserted a `self.vendordata_raw = ['#cloud-config\n']` on line 185 of DataSourceNoCloud.py | 16:15 |
falcojr | s/185/186 | 16:16 |
stevenm | has 'cloud-init' (not as a bit of software, but as a specification) ever been immortalised as an RFC or some equivalent? | 16:24 |
stevenm | because whenever someone says 'cloud-init' I've no idea if they mean the feature (of communication between host and guest) or the package | 16:25 |
Odd_Bloke | stevenm: I'm not really aware of people using it as a generic descriptor: cloud-init (the package) is present in almost every distribution, so it's usually the case that "cloud-init" is cloud-init. There have been others, cloudbase-init works on Windows too, and CoreOS' Ignition performs similar actions but in quite different ways (it runs in the initramfs, not in the booted system). | 16:48 |
Odd_Bloke | (I feel like I'm missing another one, but I may just be getting mixed up because CoreOS got acquired.) | 16:49 |
=== Savicq is now known as Saviq | ||
=== ijohnson is now known as ijohnson|lunch | ||
=== ijohnson|lunch is now known as ijohnson | ||
=== Tahvok_ is now known as Tahvok |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!