Fleeing to less worse programming languages

My career in programming has seen a number of languages. I moved from one language to another, always moving to a better language than the previous.

I started with BASIC. After a short detour with assembly language and FORTRAN, I moved to Pascal, and then C, and then C++, which was followed by Java, C#, Perl, Ruby, and Python.

My journey has paralleled the popularity of programming languages in general. We as an industry started with assembly language, then moved to COBOL and FORTRAN, followed by BASIC, Pascal, C, C++, Visual Basic, Java, and C#.

There have been other languages: PL/I, Ada, Forth, dBase, SQL, to name a few. Each had had some popularity (SQL still enjoys it).

We move from one language to another. But why do we move? Do we move away from one or move to a better language?

BASIC was a useful language, but it had definite limitations. It was interpreted, so performance was poor and there was no way to protect source code.

Pascal was compiled, so it had better performance than BASIC and you could distribute the executable and keep the source code private. Yet it was fractured in multiple distributions and there was no good way to build large systems.

Each language had its good points, but it also had limits. Moving to the next language meant moving to a language that was better in that it didn't suck as much as the former.

The difference between moving to a better language and moving away from a problematic language is not insignificant. It tells us about the future.

If people are moving away from languages because of problems, then when we arrive at programming languages that have no problems (or no significant problems) then people will stay with them. Switching programming languages has a cost, so the benefit of the new language must outweigh the effort to switch.

In other words, once we get a language that is good enough, we stop.

I'm beginning to think that Java and C# may be good enough. They are general-purpose, flexible, powerful, and standardized. Each is a living language: Microsoft actively improves C# and Oracle improves Java.

If my theory is correct, then businesses and organizations with large code bases in Java or C# will stay with them, rather than move to a "better" language such as Python or Haskell. Those languages may be better, but Java and C# are good enough.

Many companies stayed with COBOL for their financial applications. But that is perhaps not unreasonable, as COBOL is designed for financial processing and other languages are not. Therefore, a later language such as BASIC or Java is perhaps worse at processing financial transactions than COBOL.

C# and Java are built for processing web transactions. Other languages may be better at it, but they are not that much better. Expect the web to stay with those languages.

And as for cloud applications? That remains to be seen. I tend to think that C# and Java are not good languages for cloud applications, and that Haskell and Python are better (that is, less worse). Look for cloud development to use those languages.