Should bork download be removed?

[duckinator]

Basically:

• I'm not sure I have ever used bork download aside from testing the functionality.
• I literally don't know why I added it.
• It adds a lot of complexity to the codebase, for very little benefit.

I'm wondering if it makes sense to remove bork download entirely, or possibly split it out into a separate tool.

[duckinator]

@nbraud @AstraLuma what're your thoughts on this?

[AstraLuma]

It looks like it would be useful for artifact handling in CI? But

1. Very few projects publish python packages in github releases--it's just easier to use PyPI
2. I'm not sure what CI flow would involve downloading a previously-built package through non-standard channels

[duckinator]

That's why I'm thinking it might make sense to split it out into its own tool. It's useful, but in different situations than bork is useful.

[duckinator]

Independently of it being removed: as long as it exists, the definition of bork.api.DOWNLOAD_SOURCES should be moved into bork.api.download() so it doesn't affect other commands.

[duckinator]

Closing in favor of #357.