scala, kotlin, javascript, swift, dart, ...

This is a short note on programming languages (PL).

In general, I do not really care about the PL as long as it is reasonably succinct and capable to solve the problem at hand. Ideally, one PL can be used to solve different problems to amortize the cost of learning a PL over multiple projects.

Lately, everyone has defined their own programming language associated with their “platform.” Apple has Swift, google has dart (but uses java/js for other frameworks/platforms), Microsoft has C#, Sun had java, jetbrains has Kotlin. Some companies push their language as multi-platform, like Sun did, others evolve their PL to be multi-platform.

The whole PL situation is a mess. Each language is much like the other. More or less C-like with some PLs having a few more clever features than others. A company defines their own PL to control their destiny and ensure they are not dependent on another organization to evolve their platform. That makes sense strategically.

A company looking after their own interests increases the burden on the programmer. Since many PLs are similar, dart and swift are like javascript, the burden is not overly large but it does exist.

Kotlin has become popular lately for android programming. It’s a better java. Some consider a cut down version of scala. I used it for some android work and it was Ok. Adding functional programming (FP) patterns using the arrow library made it as messy, actually in my opinion, more messy, then the FP libraries of scala (and both kotlin and scala are more messy than haskell). In fact, once FP was thrown in, I felt that all the arguments against scala applied to Kotlin. And if we make it even weirder, I thought intellij’s sluggishness was noticeable and bad, a common complaint against scala and intellij together as well.

It is my opinion then that the plethora of PL is driven by corporate need and somewhat by programmer value vs programmer value first. No single PL is perfect, but many of the ones created today are just another copy-cat, so they add little value. The experience of using the language in editors can be similar as well, sometimes negative, sometimes “less-negative.”

After working with Kotlin my initial thought s that java will absorb some of its noticeable features in a few years. Perhaps kotlin will evolve just as scala is evolving. Kotline is multi-platform, more than just JVM, but that may not be enough to save it from being eaten by an evolving java.

A PL is just one dimension of a set of features that make some platforms take-off and others languish. UI frameworks have a similar pattern. I am aghast the android UI (java based). I’ve never seen such poor thinking in a UI library even if it is 15 years old, as it copies ideas from some of the original java swing libraries. The layers on top that have been designed merely added excessive complexity. Flutter is a breath of fresh air and fortunately copies the “layout/diff” approach to building UIs. Of course, some of this links back to WPF (but not exactly). They could easily evolve the android library and not create a total break that flutter represents.

It’s all a mess. I know that computer science is an ever evolving area, but you would think with all the smart people in the world working on these topics, it would be better already.