Mac Malone

Adding an Lake option for `init`/`new` that passes the option through to the Git CLI (ala `-W` for C compilers) is a good idea and would be nice to have....

@mik-jozef > And I believe this should be the default behavior as well ,,, ignores the possibility one could use a different version control system, and is also inconsistent with...

LGTM! This does not address every feature covered by #2758, so I do not think this should close that issue. Regardless, this seems like a strict improvement, so am merging...

Ah, here is the problem. Lake tests make use of the `lake new/init` to generate packages with Git repositories. That fails with this change because it is within the work...

For usability purposes, it might be best if someone other than me updated the help text and PR'd it. My level of familiarity may be disadvantageous in determine what the...

If we are not treating `StateT` as see-through, are we also not going to treat `ReaderT` as see-through? I notice that in many of these cases this is now using...

@Kha While I very much like the notion of "weak" options here, I am worried about having two different notions of weakness between Lake and Lean. "weak" in Lake means...

@kmill A rather verbose alternatively would be an `unrequired` option (or, funnily, an `optional` option ).

> Note that while this is partially inspired by Rust's `a..b`, that syntax is in fact a right-exclusive interval! But that seems like an extremely poor fit for potential openness...

Despite being generally adverse to new keywords, I think making `and` a keyword for this is a good idea. For one, Python already uses `and` as a keyword for stuff...