Frank Thomas

Results 290 comments of Frank Thomas

In the last commit I used the same artifactIds as in the play-json PR and also added the file where Scala Steward did actually update the version. It is failing...

I also tried this case with #2816 but it is also failing there.

> [...] but the key `scala` in the `ci.yml` prevents the update. If I change the key to `scala3`it works. I think we can make it work, if we add...

I suspect that calculating [`MinValue` and `MaxValue`](https://github.com/fthomas/refined/blob/f3546e2116ee42c15afff0a04293b0ee1cafa241/modules/core/shared/src/main/scala-3.0%2B/eu/timepit/refined/api/RefinedTypeOps.scala#L39-L43) of `Crazy` is the culprit.

> Would something like this work? Yes, I think so. But since this is `Or`, the `findValid` call should not be needed.

It probably makes sense to add these two instances for `Min` and `Max`. On a side note: The built-in [`GreaterEqual[N]`](https://github.com/fthomas/refined/blob/b56c90efc087472ac776daab9666f9af98df48e8/modules/core/shared/src/main/scala-3.0%2B/eu/timepit/refined/numeric.scala#L50) should be preferred over `Greater[N] Or Equal[N]`.

> So it looks like the Arrow syntax is causing this - is this intended behaviour? Not intended but expected at least. The arrow syntax is an implicit conversion and...

Try `Size[Equal[n]]`.

> Validator is still missing. Right. You're trying to check if the string representation of an `Int` has `n` digits. We don't have a predicate for that. > I also...

> I didn't really got what it is about in macro_pitfalls. My recommendation is to not use refined's macros at all. They cause problems like above and are not available...