Rename ghost'' to ghostWith

[Zalastax]

Keeps ghost'' as an alias.
Closes #958

[mindofmatthew]

Looking through the issues, I think this actually closes #958?

[Zalastax]

Good catch! I'll update the commit to fix this typo.

[mindofmatthew]

I don't have approval permission, but I think this is worth merging. In general, I think v1.10 could benefit from cleaning up and standardizing the API a bit. Tidal currently exports 800 functions and 1300 params, only a fraction of which are documented. I'd propose the following guidelines for this process:

• Retain all old names as aliases for now
• Prefer camelCase version in function definitions and documentation (with lowercase aliases in code)
• Move away from "prime" function notation (func' or func'') (as it doesn't communicate anything about the function itself and it can't be represented in Strudel)
• Adopt a more consistent vocabulary, eg:
  • funcBy generally means that an operation has an extra numeric argument, specifying how much or by what probability an operation is applied
  • funcWith generally means that an operation has an extra functional argument, specifying a function that further modifies the added/generated/modified pattern
• Maintain parity with Strudel
• Identify which functions are internal and explicitly exclude them from export lists

Anything I'm missing? I can do a first pass and drop a list of specific suggestions in an issue for folks to comment on, unless there's some other way we should approach this.

[yaxu]

Ok merged!

@mindofmatthew thanks that's well thought through, I agree with every one of those points!

The complication is that I've put a lot of work into #947 on the cycseq branch, which is a major reorganisation towards Tidal version 2.0, based on the Tidal 'remake' (that went into vortex and strudel). Would you be up for looking at that instead? I haven't looked at it in a little while, it's not complete but it's passing most tests. It could be a good basis for standardisation.

[mindofmatthew]

@yaxu Yes, I can definitely focus my attention on the cycseq branch, though I'd be inclined to submit parallel PRs against main if any particularly useful changes come up. I agree that 2.0 is the place for a broad reorganization.