New Maintainer proposal and suggested plan for moving forward

[joemoore]

TL;DR: Unless someone stops me I'm going to take this thing over, modernize it, get it sane, strip out the cruft, make almost everything optional, all for the greater good. Not everyone will like all of the changes, but most everyone will agree they make sense. Once updated we can reimplement a healthy PR culture. I will communicate this proposal in as many places as possibly but will assume silence is acceptance.

---

Nobody is maintaining this tool. It has a ton of forks and not a lot of people contributing back.

People still use and rely on this tool, especially people from the Pivotal heritage, such as VMware employees, Pivotal alumni, current clients, former clients, partners, friends, family, loved ones.

I hereby nominate myself to maintain this tool for now in an effort to modernize it, bring sanity, and draw a line-in-the-sand for moving forward.

My qualifications:

• I have worked on and used this tool and all of it's predecessors for the past 15 years
• I'm a known quantity by this tool's primary audience: people from the Pivotal heritage
• I'm fair and the above mentioned community tends to trust my judgement
• I'm not afraid to make hard decisions and make some people unhappy on behalf of the greater good
• I have the time, tools, and enthusiasm

Here is my current plan:

• Declare PR bankruptcy
• Put a pause on design-by-committee and implement DESIGN BY JOE MOORE THE DECIDER OF ALL THINGS DAMNIT
• (But just for now I promise. We have to make some progress as fast as possible and eliminate lots of cruft and get to a sane baseline in order to move forward. Plus I'll constantly communicate.)
• That means personal opinionated changed in current PRs will mostly be ignored. If you want a different default terminal font or editor line length you'll have to do that yourself because the greater good.
• What's the difference between a sane default and a personal opinion? I DECIDE. Again, just for now. We can duke it out in PRs later.

I'll take inspiration from current PRs and create a series of new PRs with a goal of:

• Making certain must-have changes, such as zsh by default.
• Remove anything that is flat-out broken or dead
• Decompose the current installer scripts into their contingent parts. For example, the fact that git.sh installs vim is bonkers. Again, current PRs have some good ideas.
• Make as much as possible optional and opt-in. This likely means commenting most things out and providing instructions to uncomment what you want. No fancy choose-your-own-adventure set of command line args. We're developers. We basically live by commenting things in and out and running it again anyway.

Am I going to merge in my own PRs? Yes, I probably will. Again, I'll constantly communicate. Again, once we're at a new sane baseline we can have a healthy PR culture.

I have an OS X VM running in Fusion on my mac plus an extra loaner mac to test stuff. I can also look at some CI solutions, especially if GitHub Actions supports an OS X runner.

Let me know what you think.

[rkawala]

I second. I'm sure you'll do a wonderful job. I'm grateful that you're willing to take it on.

[ericadohring]

What a treat of a read 😆

• "DESIGN BY JOE MOORE THE DECIDER OF ALL THINGS DAMNIT"
• "I DECIDE"

Re: CI Solutions - Circle has a Mac executor, but I believe the last time I looked into it, it cost $ and I didn't know if I'd get buy in to expense a using a non-Tanzu product for a Tanzu thing.

Also ... sadly the Circle Mac executor comes with brew already installed. Not the worst thing, but also not a perfectly pristine test environment 🧽

Lastly comes the question - is maintaining this repo a futile task? Github recently released a blog post (which I'm sadly not finding quickly) about their workstation setup. It's more containerized. But that could also be specifically to their application, of which this repository does not deal with any 1 application ... anyway, I personally love and support WS setup, but just wanted to throw that out there /long winded comment.

Thanks for taking this over, Joe!

[bastiandoetsch]

Yeah go ahead with it. And thanks for picking this up, I feel it is worthwhile the effort. Feel free to cherry pick or merge my stuff, or not - I'm completely fine with you taking the decisions on upstream.

[joemoore]

Lastly comes the question - is maintaining this repo a futile task? Github recently released a blog post (which I'm sadly not finding quickly) about their workstation setup. It's more containerized. But that could also be specifically to their application, of which this repository does not deal with any 1 application ... anyway, I personally love and support WS setup, but just wanted to throw that out there /long winded comment.

This is a very interesting and valid question. I think that eventually we'll all be working primarily in containers. I'd still like to get this tool in better shape and then solicit volunteers to do a kind of bake-off.