Time for a monorepo

[tivac]

We keep on running into issues with things not being published and the repo root is starting to get messy. I think it's time for a monorepo.

@pygy, @isiahmeadows, @lhorie, anyone feel strongly for or against this plan?

Work has already started here: https://github.com/tivac/mithril.js/tree/monorepo

[pygy]

I'm for it, but I'd rather wait for #1592 to be merged in if you don't mind (unless git is smart enough to track moving files when rebasing).

[tivac]

I'm not in a huge rush, happy to manually merge stuff as it comes in.

[dead-claudia]

I'm for it, since we're already half-way there now. Just a few nits on that repo:

1. Why is mithril-querystring separate from mithril? It's not already published independently, and it doesn't have much use case outside Mithril. (Most people looking for a querystring module would just use qs).
2. You forgot to list mithril-stream as a dependency of mithril.
3. The tests module seems pretty pointless IMHO. We should probably just move the lone test file to mithril/api/tests, where it would make more sense.

[tivac]

Both mithril and test-utils depend on mithril-querystring. I could introduce a bunch of graph cycles by making test-utils and mithril depend on each other but it makes file explorers and editors freak out. I'll probably just make test-utils depend on a 3rd party parser instead long-term, this was a quick pass to get tests running.

[pygy]

External dependencies will break the tests/index.html pages.

I thought that mithril did not depend onmithril-stream... does it?

[tivac]

It ships stream currently, at least.

https://unpkg.com/[email protected]/stream/stream

I'd be totally fine with ceasing that practice though, I don't think that is a necessary part of mithril core to ship with a npm install.

[dead-claudia]

@tivac To clarify, I'm referring to this file pointing to a directory that doesn't exist. If we decide to unship it (I'm neutral on it, since I almost never use it personally), then my nit is moot.