[x0x0]

> It's not that I actually believe most of these things are "useless". But rather that my mind can't draw a good problem scenario for them. So from my POV they look like solutions looking for a problem. Maybe my experience just has no good frame of reference for those things.

For lots of things your frame of reference should probably be sth like replacing very expensive and very limited engineering time with inexpensive tooling. That could be inferior. But I'm far more eng time limited than I am cash limited. Or, particularly for deploys, (1) blue-green deploys are great; and (2) it's easier to hire eng that don't understand how to use rsync or capistrano than it is to hire those that do. Just like it's easier to use github actions to make it very hard to deploy a branch w/ failing tests than it is to just rely on everyone running tests. (Which, in a perfect world, you could do... but here we are.)