I dont think 15min is too bad. we probably should not spend too much time for the nginx spec files. but converting more templates from erb to epp in our modules and adding a global cache in rspec-puppet-facts is probably a good idea

ewoud: any chance you can revisit https://github.com/voxpupuli/rspec-puppet-facts/pull/124

bastelfreak: https://github.com/voxpupuli/puppet-nginx/pull/1565 is one way to speed things up

you can see it took 7.6 seconds to load and that's the mostly the rspec-puppet-facts overhead

and I don't see myself having time for that soon :(

ewoud: now that we have money we can just outsource stuff like that to people that are smarter!

or people that just have time

🤔

yuri just volunteered

*yury

:-D

we could also offer a bounty

I can spend some time migrating erb to epp maybe

perfect <3

I got a well-paid full-time job.. which is nice but now I don’t have enough time for all the interesting things :)

I wonder if someone could write a basic erb2epp script

But anyway I’ll try to have a look on templates in the nginx module this weekend

perhaps even extract all variables used and define them automatically with the data type Any

interesting idea

Temple:ERB:Parser can parse an ERB file, so that would be a start

ewoud: :D

not a script, but still scary how good the response is

indeed

so I took a quick stab at it and https://gist.github.com/ekohl/28443e6ba178c931bb69285e976568eb is what I came up with

looks like it's generating some bad things though :P

but if you took a ruby parser to the dynamic parts and rewrite it there in an AST you could maybe do cooler things

bastelfreak: I did a trivial rebase, which turned out to apply cleanly

awesome, thanks

@Yury Bushmelev maybe you can give that a try

if we say that asking for version x.y.z should strictly return facter versions <= x.y.z then I think it's safe to merge, but I think it implies we must update ext/puppet_agent_components.json quite regularly