Remove MultiProgress::join() based on #231

[aj-bagwell]

This change moves drawing of MultiProgress from the join() thread to the child ProgressBar updates. fixes #125, #33 and lay groud work for work on #205

It is backwards-incompatible.

It is based on the work of @marienz in #231 more discussion and detail on that pull request.

[djc]

Thank you for working on this! In order to make sure I can review this closely, I'm asking for some changes to the commit history presented in this PR. I usually use git's interactive rebase to do that, but if you're unfamiliar with that stuff I can also help split it up more. The goal is to have a few commits that each independently could build and pass the tests.

Also, I think the reference to #116 in the second commit's message is wrong, can you fix that up? It'd be nice if you could shorten the first line of that message to fit within 70-80 chars while you're in there, and a high level description of what problem the commit is trying to fix and the strategy for doing so.

[boc-tothefuture]

I don't know enough about indicatif to know why it causes the defect, but the current PR does not respect when a progress bar is hidden.

        let mp = MultiProgress::new();
        let hidden_bar = ProgressBar::hidden();
        debug!("Bar Hidden?: {}", hidden_bar.is_hidden());
        let bar_after_add = mp.add(hidden_bar);
        debug!("Hidden after add?: {}", bar_after_add.is_hidden());

Corresponding output:

Bar Hidden?: true
Hidden after add?: false

[mibac138]

@boc-tothefuture are you sure this PR actually changed this behavior? See the docs for set_draw_target which note this case explicitly in the current version.

[boc-tothefuture]

@boc-tothefuture are you sure this PR actually changed this behavior? See the docs for set_draw_target which note this case explicitly in the current version.

You are correct, the same behavior exists in the currently released version. I did not think I was experiencing that until I switched. My apologies.

[djc]

Except for the typos, this is looking pretty good! Thanks for working through my comments.

[aj-bagwell]

Typos fixed also fixed the failing CI

[djc]

I just pushed another test that calls MultiProgress::join(), would you mind rebasing on main and pushing that once more? I can merge it after that.

[aj-bagwell]

Rebased on the the latest main (ignore the moment of stupidity where I rebase it on master instead).

Thank for all your help an patience with this.