https://www.puppet.com/community logo
Join Slack
Powered by
# voxpupuli
  • v
    we once had a hack to read a yaml file in, which method did that use? that worked fine back in the days
  • v
    I'm really confused because I'm sure I used this code before and it worked
  • v
    but now I'm wondering if it really did or if we just weren't using it
  • v
    so my current theory is that Puppet/Facter doesn't care about symbols/strings and only if you use another custom fact in your custom fact you hit this
  • v
    right, I'll think more about this after lunch
  • v
    ok, I'm reasonably sure now that the code I wrote then for the systemd custom fact never worked and I never needed it before
  • v
    I know a lot of modules use it,but we also often set it explicitly in the let facts block
  • v
    maybe we always set is there
  • v
    well, this code isn't used in VP (yet)
  • v
    https://github.com/voxpupuli/modulesync_config/pull/693 would do that but I never got around to pushing for it
  • v
    and recently I did migrate theforeman modules but in the one module that does use it CI was turned of (thanks GHA for disabling workflows with cron after 60 days)
  • v
    mhm
  • v
    we deleted the yaml file that mocked facts, I thought we replaced it with add_mocked_facts
  • v
  • v
    yes, the nuance is that you end up with (in Ruby) facts[:osfamily] and facts['my_custom_fact']
  • v
    ah we only added three facts, not the systemd one
  • v
    and my code checks facts[:my_custom_fact] - which fails
  • v
    I think it may be time for a rspec-puppet-facts 3.0 which cleans this up, but not today
  • v
    https://github.com/voxpupuli/voxpupuli-test/pull/81 I think that documents today's journey
  • v
    approved it
  • v
  • b
    @Alex Fisher do you want to do new releases for the puppet-lint plugins you updated?
  • a
    yeah. Will raise PRs for new releases.
  • v
    It seems python 6.3 should have been v7, since it introduces incompatible change. It starts managing venv package and removes it by default.
  • v
    also, this logic is questionable
    Copy code
    unless $facts['os']['name'] == 'CentOS' {
    imho, should be driven by hiera
  • b
    mhm, was there a PR with the wrong label?
  • v
    enhancement
  • l
    gotta love that redhat has removed lsb_release support
    🙈 1
  • v
    what was the point of it anyway?
  • l
    it provided a standard way to get os info
1...104105106...648Latest