Hey folks,

I'm working at a company with dozens of Scala developers, and we recently started adopting Scala 3. So far, so good—except for the mixing of Scala 2 and Scala 3 syntax in Scala 3 codebases. We assembled a small team who "jumped into the water" and wrote a brand new service in Scala 3. The service works well, but the code ended up being a mess of implicits and givens, classic and braceless syntax, as well as both the old and new conditional syntax.

While I understand that this mix can be useful during the transitional phase from Scala 2 to Scala 3, it’s a disaster for a new project. Scala isn’t the simplest language to learn, and having both Scala 2 and Scala 3 syntax in the same codebase only adds to its complexity.

Fortunately, we discovered two compiler flags—-old-syntax and -no-indent—that help maintain consistency and reduce the syntax variety in our code. However, some features, such as implicits (which have been replaced in Scala 3 by givens and extension methods), cannot be disabled or phased out via a compiler flag.

For someone starting a new project, it makes sense to write pure Scala 3 code. I believe the compiler should offer a "strict" mode that disables backward-compatible Scala 2 syntax entirely.

What do you think?