Closed benbhall closed 3 years ago
In favour of a allow list approach - I'd appreciate words like just, simple and easily being flagged in my writing to check if I am being condescending in the context.
It's a shame the guidance with the Alex warnings is not more detailed. The file in the retext-equality repo, for example, explains why just can be inconsiderate quite succinctly:
Not everything is as easy as you might think. And if it isn’t easy for the reader, it can hurt.
Following on from #109.
Alex.js still has fairly limited control but I don't think it is worth diving into other tooling like textlint, which will have it's own issues.
We can avoid warnings such as 'UK' by setting:
"profanitySureness": 2
.Just need to decide on a workflow:
Deny list
Set only
deny
but risk being very limited in what we catch e.g.Note those are rules, not words. See https://github.com/retextjs/retext-equality/blob/main/rules.md.
Allow list
We set only
allow
. Note deny and allow and mutually exclusive. This is my preference, although heavier upfront work on existing sources. This means all possible warnings that we might not consider are flagged. The author can either make the suggested changes or add to theallow
list.Allowed rules can be managed in different ways depending on situation:
Inline through comments: https://github.com/get-alex/alex#control e.g.
<!--alex disable dad-mom-->
In our
package.json
e.g.