Handwringing about Scala

There has been a flurry of blogs and email threads with hand-wringing about the fact that Scala hasn't yet achieved world domination.

Scala is too complicated. Sure, it has its complicated corners, but complexity hasn't stopped C++ from being #3 in the TIOBE ranking for over ten years.
Scala has no killer app. That's true—it doesn't. What was the killer app of C++ again? I forgot.
Scala tooling is atrocious. Is it? I think it's now where Java was around 2002 (seven years after its first release). Think back—that's when the first usable versions of Eclipse and NetBeans appeared. People who complained about their shortcomings were told to try IntelliJ :-) And the Scala tooling is improving at a pretty nice rate. By the way, I don't think PHP, Python, Perl, Lua or Ruby tooling is all that great either, but that hasn't stopped those languages from being #4, 8, 9, 10, and 12 on the TIOBE scale.
Unlike Ruby, Scala doesn't tell people one true way of doing their work. Well, Ruby doesn't do that either. Rails does. And PHP? Don't get me going.

I think Scala is going to grow steadily and undramatically because it's such a natural progression from Java. If you are a Java programmer and want to learn something new that at the same time expands your horizon and makes you more productive, what are your choices?

C#. That's not really going to expand your horizon
Haskell. By all means, but it doesn't help you a lot with your day-to-day work
Racket. Ditto
Lua. Objective C. PHP. Those might contract your horizon
Ruby. A contender
Scala. The easy choice. Interoperable with Java. Less boilerplate. Ease into functional programming. What's not to love?

That's what I see as the second phase of adoption (after the first one—by the language enthusiasts who currently dominate the Scala scene). Of those individual people, more and more will release toolkits and frameworks in Scala, not Java. The third phase will be, as with all successful languages, consolidation, compatibility, and stagnation.