| sbraz | hello, why isn't https://github.com/canonical/cloud-init/commit/076b51b362402d821414ee329a90c6ecd5b8aa7a included in tag 22.4.2 (and therefore Debian 12) although it is included in e.g. 24.1.4? | 10:12 |
|---|---|---|
| -ubottu:#cloud-init- Commit 076b51b in canonical/cloud-init "network: Deprecate gateway{4,6} keys in network config v2 (#1794)" | 10:12 | |
| meena | sbraz: that's very odd | 12:25 |
| meena | sbraz: would you mind reporting that as issue? | 12:26 |
| meena | sbraz: the problem might be that it was included in the 22.4.2-0ubuntu* tags, and someone forgot to pull it into the actual release tags, so if debian builds from 22.4.2, and ubuntu builds from 22.4.2-0ubuntu2, you'll have drift between their packages | 12:33 |
| sbraz | meena: yeah i noticed that it was in the ubuntu tags, it is indeed weird, i assume that debian builds from the non-ubuntu package because the patch is not present on Debizn 12 | 12:37 |
| sbraz | i'll create the issue immediately | 12:37 |
| meena | basically, the only way to fix that is to issue a 22.4.3 | 12:38 |
| meena | or if you import the patch… as patch | 12:38 |
| meena | or if you update to a newer version | 12:38 |
| sbraz | meena: of course, hopefully Debian would upgrade to 22.4.3 if/when you do this | 12:39 |
| meena | the way debian releases work, there's no chance | 12:40 |
| sbraz | meena: not minor enough to warrant inclusion in a stable release? | 12:41 |
| meena | argh | 12:41 |
| meena | sorry | 12:41 |
| meena | I misread 22.4.3 as… a different number | 12:41 |
| * meena has something like dyslexia/dyscalculia but not quite | 12:42 | |
| sbraz | meena: oh actually you're right, this commit was never included in 22.x tags, except the Ubuntu ones, 24.1.4 > 22.4.2 | 12:46 |
| falcojr | sbraz and meena: this first appeared in 23.1 | 12:51 |
| meena | falcojr: but github says it's in earlier -ubuntu tags | 12:52 |
| meena | anyway, back to trying to figure out why this damn test is failing | 12:53 |
| falcojr | if you're talking about `22.4.2-0ubuntu3` and `22.4.2-0ubuntu2`, those were uploads to ubuntu/devel which is the pre-release version | 12:54 |
| falcojr | we've since renamed how we upload to ubuntu/devel to more accurately represent that it's really a prerelease of the next version rather than a supplement to the previous one | 12:54 |
| meena | aaah | 12:57 |
| meena | now i get it | 12:57 |
| meena | y'all, I need some hand-holding from someone who knows what they're doing with Python tests, because I have no idea | 15:33 |
| meena | hrm | 15:47 |
| meena | maybe I just don't know how to read | 15:48 |
| meena | I should build a VM with that code, just in case I changed anything since trying to add unittests | 15:59 |
| meena | I'll give https://github.com/canonical/cloud-init/pull/5169 another testrun on AWS, and if it still works, it's actually ready for review | 16:04 |
| -ubottu:#cloud-init- Pull 5169 in canonical/cloud-init "growpart: Fix behaviour for ZFS datasets" [Open] | 16:04 | |
| holman | meena: sounds like progress :-) | 17:32 |
| meena | holman: i have now tested it, and it still works, so i think it's more than progress | 18:34 |
| meena | Although I'm not happy with those tests, they all feel fake | 18:34 |
| holman | Nice! | 20:16 |
| holman | Yeah an integration test would go a long ways with this kind of thing | 20:16 |
| holman | I thought of a way to make the current integration tests not be tied to a single cloud | 20:17 |
| holman | Still hacky, but in bootcmd we can create a sparse file-backed block device and do whatever operations we want in our other modules | 20:19 |
| meena | oh, makefs -t zfs works on 14.0! | 20:47 |
| meena | (I thought it was only in CURRENT) | 20:48 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!