[Improvement] Simplify Interations

[guidiego]

Resume

I think that improve docs and change something on command interface could improve a lot of experience with kube-render, but here some tips that could help :)

Merge -u and -t

Basically, we could identify if it is a github or local path, more improvement tips here (#17)

Remove -w

Once that we have -c to add yml context files we don't need to specify a "workdir", we could only pass a full path on -c and it's easier.

Main call

If you give a kube-render it throws an error, we could only call --help instead.

[andrewsmedina]

I believe that we should figure out a way to improve kube-render without to break the CLI interface. So, how some of these proposals will break the interface I'm will think twice before to do it.

But if we believe that this improvements will improve a lot the kube-render usage, my suggestion is do it in a way that does not break de CLI:

So, for the suggestion to remove -w, we should deprecated this parameter for a long time before to remove it.

For merge -u and -t we can to it creating another parameter and deprecated these parameters.

And for show --help on errors, I think that is not a good idea do always do it. But it can be done when the error is related to CLI commands or parameters.

[guidiego]

@andrewsmedina let's go!

About deprecated params, +1. O think workdir an "overflag" as I explain. About merge, I like the idea to create maybe a --source -s and deprecated the others.

I want to throw --help only in the case that I call kube-render without ANY flag or parameter 🚀

[guidiego]

@andrewsmedina Another deprecated approach is: Create a second version os kube-render, and maintaining the new one after a period of time.