Albert Meltzer

> The `q"..$mods def $name[..$tparams](...$paramss): $tpe"` don't work, @kitbellew is that what you meant by quasiquotes ? How can I fix it ? yes, that's what i meant. i don't...

@Sporarum fyi, the recommendation is for you to preserve the old parsing and create new top-level trees and put the new parsing under a dialect flag: https://github.com/scalameta/scalameta/issues/405#issuecomment-1237307603

thanks, @tgodzik . @Sporarum i would also like to point out that naming (both here and in SIP) is a bit confusing and broad. this is not "clause interleaving" but...

> Now for dealing with arguments that span several lines themselves in binpack mode, the rule is quite simple as well: > > * As soon as an argument spans...

@ekrich recommend trying 3.0.0-rc7 with the following additional configuration options: ``` fileOverride { "glob:**/scala3/**" { runner.dialect = scala3 } } optIn.breakChainOnFirstMethodDot = false # preserve some overflow newlines.avoidForSimpleOverflow = [tooLong,...

> [diff.txt](https://github.com/scala-js/scala-js/files/7367549/diff.txt) I do see some strange indentation. e.g. OptimizerCore.scala line 261 or line 49089 in this diff file. 4 for case continuation, 4 for tuple (`(LongType | ..., ClassType(...))`)...

@ekrich i thought this repo had a desire for this: https://scalameta.org/scalafmt/docs/configuration.html#inserting-braces

> @kitbellew Should this be the case since `binpack.literalArgumentLists` default is true? > > ```scala > --- a/Users/eric/workspace/scala-js/compiler/src/test/scala/org/scalajs/nscplugin/test/JSGlobalScopeTest.scala > +++ b/Users/eric/workspace/scala-js/compiler/src/test/scala/org/scalajs/nscplugin/test/JSGlobalScopeTest.scala > @@ -331,9 +331,49 @@ > final val ReservedJSIdentifierNames:...

> The wrapping of args in docs also creates quite a bit of extra diff - might look a little better? > > ```scala > + * @param args >...

> This kind of change is annoying, but I understand it's going to be _necessary_ for indentation-based users, so I guess scalafmt won't invest in supporting our previous format: >...