Extension talk:WikiLambda/Discovery needs

One thing that worked well on Wikidata was to not prio them by popularity alone, but rank by best fit to conditions, and if there are not that many hits, relax the conditions and extend the list.

Yeah, possibly. I'm thinking that the filters will be sufficiently strong that people will normally want prioritisation, but will adjust wording to make that clearer.

Thanks, yes, that's great!

I still hope we can get away without having explicit subtypes, but one other example could be:

• As a function user, show me a list of Functions which take exactly a list of strings
• As a function user, show me a list of Functions which take exactly a list of anything
• As a function user, show me a list of Functions which take exactly a pair of string and boolean
• As a function user, show me a list of Functions which take exactly a pair of string and something else
• ...

So basically generic types where the argument might be not given (i.e. anything). Those are not defined as subtypes (i.e. we don't say "string16 is a subtype of string") but rather these are structurally necessary subtypes (list of string is a subtype of list of anything) - which makes the hierarchy static and predictable.

Good idea; how's that?

Perfect, thank you!