When using Scala for the first time in the IDE (Idea), I noticed that autocompletion is noticeably slower, and then when coding java. Although some of the slowdowns may be related to the relative immaturity of the Scala tool ecosystem at the time, I suspect that some of this slowdown may be an integral property of the algorithmic complexity of code analysis requiring type inference.
java:
MyType type; type.doSomething() //Class of type already known
scala:
val type = new MyType; type.doSomething() //Class of type must be inferred or cached
Although languages ββthat have type output are much more concise (and therefore easier to read), does this come at the expense of slower tooling? Is there an inherent compromise?
source share