RFC exclude_author_names -> alternative_author_names

[tiborsimko]

The current configuration option to accept alternative author names while checking the completeness of the AUTHORS file is called exclude_author_names (the meaning being to "exclude" these false positives from being reported). This may appear harsh, as if we were excluding some contributing authors from getting credit or something...

What about renaming it to something nicer? Say alternative_author_names.

(stemmed from inveniosoftware/cookiecutter-invenio-module#60 (comment))

[jirikuncar]

or simple author_aliases?

[tiborsimko]

Hmm, when talking about "alias", people may expect to see the target value specified, as in alias "Johnny Doe" -> "John Doe", while currently we specify only the names to ignore...

[jirikuncar]

Can we use an "empty" alias to trick it? Normally, all committers should be in AUTHORS.rst.

[tiborsimko]

I was just wondering about the appropriate naming semantics if we use the word "alias" to represent something that is a "list".

To make it more explicit, we currently have:

exclude_author_names:
- David Caro <[email protected]>
- Jiri Kuncar <[email protected]>
- Mateusz Susik <[email protected]>
- Yoan Blanc <[email protected]>

The proposed alternative_author_names seems nicer:

alternative_author_names:
- David Caro <[email protected]>
- Jiri Kuncar <[email protected]>
- Mateusz Susik <[email protected]>
- Yoan Blanc <[email protected]>

If we use author_aliases:

author_aliases:
- David Caro <[email protected]>
- Jiri Kuncar <[email protected]>
- Mateusz Susik <[email protected]>
- Yoan Blanc <[email protected]>

then it may not look right if people expected to see target values following the alias foo bar analogy, which would lead to something like:

author_aliases:
  - David Caro <[email protected]>
     - David Caro <[email protected]>
  [...]

which would be an overkill.

This is why I'd rather prefer a "list" like name rather than an "alias" like name.