Keean Schupke

Results 326 comments of Keean Schupke

> We avoid that but add the boilerplate back in other places. So isn't necessarily a net gain in brevity. It also loses some local reasoning. Callbacks are not local...

> Generally wrapping interacts badly because it becomes deeply nested!!! It is still better than the alternative of a huge multi-line type signature all in one block.

@shelby3 looking at the direction you want to go, we may be better off ditching quantified types and type-classes altogether and going for a full intersection and union type system...

@shelby3 I feel you are making statements about a style you have no experience of coding, so I am not sure how much weight to give your objections.

> I haven't made any decision. I am trying to analyze the trade-offs. Intersection types allow functions to be like a type-class where a single function can have different signature...

> The context of my point was MyCallback and MyCallback2 both having the same structural type Why would you declare two of them then? Thats like saying `add` and `add2`...

The 'Product' type constructor is parsed and printed as a type-list in angle-brackets, like this: ``` -> ```

After using the type syntax for a while, I now think that the angle brackets '' look overused, being part of the arrow syntax as well. I have swapped to...

I have been looking at PureScripts use of row types, and row polymorphism, and I think this is the way I want to go for effects and union types (at...

@shelby3 JavaScript Objects can be represented by a row type like this ``` forall E . { label1 : Type1, label2 : Type2 | E } ``` Its basically a...