[14:29] <meena> falcojr: i tried reverting rharper's changes to see if i can pin-point the failure, but I can barely get a sensible debugger session going on
[14:30] <meena> I feel like our test suite is a bit too much for (i)pdb
[14:54] <holman> meena: not sure if this helps: my workflow is to drop `import pdb;pdb.set_trace()` into any test that I'm inspecting.
[14:54] <meena> holman: done that… it doesn't stop there
[14:54] <holman> it doesn't stop at the breakpoint? 
[14:55] <meena> holman: just flies right past it… and when i add -s to the invocation, it just straight up becomes useless
[14:56] <holman> it should stop when it hits the `pdb.set_trace()` line
[14:56] <falcojr> meena: did you apply the diff I put on the PR? It doesn't just revert Ryan's change
[14:57] <meena> falcojr: no, i was staring at the screen half day long
[14:57]  * falcojr not sure what that means :P
[15:01] <meena> falcojr: I was too busy staring at ipython/ipdb failures and trying to figure out how to get a useful debugging experience, that I didn't look at what github is doing
[15:08] <falcojr> also, cloud-init does mess with stdin, which can mess with debuggers, but it shouldn't affect most tests https://github.com/canonical/cloud-init/blob/main/cloudinit/cmd/main.py#L330
[15:21] <meena> falcojr: I'll try your changes out in a bit…
[18:23] <meena> whoohooo! I have an aarch64 FreeBSD 13.1 running in a local VM
[18:26] <meena> oh, but I don't have a custom aarch64 cloud-init yet…
[18:31] <meena> falcojr: doesn't work
[18:31] <Minghe> Hello folks, when I try to run "tox -e do_format" to test my changes, I am getting errors:  subprocess.CalledProcessError: Command '['git', 'describe', '--abbrev=8', '--match=[0-9]*']' returned non-zero exit status 128.
[18:31] <Minghe> Anyone knows how to solve this?
[18:33] <meena> falcojr: https://github.com/igalic/cloud-init/tree/better-fix/cc_ntp_bsd
[18:35] <falcojr> Minghe: Do you have git installed? Does git give you any other problems?
[18:37] <Minghe> yes I am on ubuntu and have git installed. Previously, I had an error about git branch --show-current and I upgraded git to version 2.38.0 and solved that one
[18:37] <Minghe> now I am getting error on '['git', 'describe', '--abbrev=8', '--match=[0-9]*']
[18:38] <falcojr> Minghe: what happens if you run that git command directly on the cli?
[18:40] <meena> falcojr: i get a mix now: E               AssertionError: expected call not found.
[18:40] <meena> E               Expected: subp(['systemctl', 'reload-or-restart', 'ntp'], capture=True)
[18:40] <meena> E               Actual: not called.
[18:40] <meena> tests/unittests/config/test_cc_ntp.py:492: AssertionError
[18:40] <falcojr> meena: Can you apply my patch directly? I still see some significant differences
[18:40] <falcojr> I.e., save the diff to a file an git apply <file>
[18:41] <falcojr> I just pushed it here too: https://github.com/TheRealFalcon/cloud-init/tree/pr_ntp
[18:41] <meena> that's with rharper's one completely removed (git rebase -i HEAD~6 and remove his patch, and my isort patch)
[18:41] <falcojr> I'm mocking subp differently
[18:42] <falcojr> mock the function rather than the module
[18:45] <Minghe> I am getting error: fatal: No names found, cannot describe anything if I run git describe
[18:46] <meena> falcojr: oh, right, forgot about changing that in the assert
[18:48] <meena> falcojr: okay, so your patch does, indeed work… why?
[18:48] <meena> (i mean, i could ask the same about rharper's, but yeah)
[18:55] <falcojr> meena: Not entirely sure. I'm fuzzy on the details of packages and importing, but in general I don't mock entire modules 
[18:56] <falcojr> Minghe: Have you cloned the entire repo? A lot of cloud-init's tooling won't work with shallow checkouts that only give you HEAD. You need the full history
[18:58] <Minghe> I see looks like my cloned fork does not have any tags or other branches
[20:26] <meena> https://im.eena.me/uploads/ac588cbec33ad836/image.png Co-authorship with rharper ended, now falcojr is my co-author.
[20:36] <meena> that is to say: falcojr I've redacted rharper's change (completely) and pushed yours
[21:05] <meena> gonna try and test OpenBSD now… wish me luck…
[21:15] <minimal> sam_: have you ever used cloud-init on Gentoo when it has openrc rather than systemd in use?
[21:33] <holman> minimal: I have a branch I need to revisit when I find time that has a couple of fixes for openrc/cloud-init
[21:37] <meena> minimal: new set of changes is much simpler
[22:19] <minimal> holman: I have changes regarding openrc that I've already made locally for Alpine and was waiting for meena's stuff to be merged before submitted them, but was wondering about Gentoo as changes I've made to manage_server in alpine.py would likely also be applicable for Gentoo with OpenRC
[22:24] <minimal> s/manage_server/manage_service/