@chas @feld I’ll just reply in line not any specific direction.

Ansible is good for what it does. I won’t deny but it’s a pile of shit if you really need to get things done fast.

Saltstack it’s like ansible with afterburners. That’s a fact. They didn’t do well with pull requests. There’s one outstanding I did for tar exclude that’s rotted since 2018. Ended up I just started making my own modules and states, custom utils and all that. Reimplementing many half ass items they did.

Between both they didn’t work towards a usable end product as they ultimately became to sell a product. That product wasn’t what I used so I was off in the wind.

I’ve been doing automation since 2015 at least.

Now let’s dig into the details. Why was salt so concerned about modification to their code base? Well go look at it. Tell me what rational human could handle that code base. It’s duplicated code like CRAZY and the amount of changes to not be breaking? Means that it became their modules are stupid.

Here I’ll just replace it. That’s what it became was not using salt as it’s documented. But to use the salt engine to run my code. That made me maintain so much code. It was really a lot of work for very little return after using chef. I wasted my life lol.

Then they sold to VMware and it shit the bed. There is no value there.

At any rate, I’ve got complete FreeBSD management tools on saltstack that are useless now. It’s not worth my time. Not to mention python really does suck.

The day I stopped using salt was the day I stopped using python. That was the only reason I was using it.

Blah is all I can say.

Between all these and then using chef? I wanted to write my own solution but chef works. Randos develop for it as they dare to care. Unlike salt or ansible you don’t see the same level or commitment.

I made this and then tried to talk about it. Apparently it was “so big” that it would require tons of review. Just based on the lines of code. Well that’s fucking stupid. What does lines of code have to do with something that doesn’t really touch anything else?

https://github.com/SlicerDicer/SaltThings

Wow too 3 years ago damn. Well that should give the clue that I gave up with any kind of contribution.

Someday I’ll deal with it and make something that really fits my need. Till then? Chef is a wonderful toolkit for getting things done. Way less hassle than trying to bend salt to my will. I gave up on ansible in 2017.

As to DeHann? Well he tried to do other automation solutions that were designed better. That was opsmops but it never took off. I had at length discussions about this with him actually. He did try to but inertia is always a hell of a problem.