[pedasmith]

I have to both agree and disagree here.

I love the idea of a language that has the absolute fewest number of decisions between "I need a little program" and being able to share the program with a friend. The current trend where there's a seemingly bazillion steps before you can write a trivial utility is unfortunate.

However, the microcomputer BASICs had big usability problems. Us modern programmers are so used to block oriented languages that we assume every language just had them. For example, in C you can replace any statement with a list of statements enclosed in { }. In Python where blocks are indented statements, or Pascal where you can make a block of statements with a BEGIN/END.

But many microcomputer BASICs don't have that at all. The IFs often don't have the concept of an ELSE, and often the only thing you can do from an IF is a GOTO.

The end result is a mess of jumps where the programmer is "emulating" real blocks with GOTOs. I've had the recent fun of transcribing programs from a bunch of different early BASIC languages. The complaint that BASIC leads to spaghetti code is real. There are bazillions of GOTOs, and it's a real drain on your mental processing to keep them all straight.

[pjmlp]

Sure, but BASIC doesn't need to be the 8 bit versions or GW.

Something like Turbo Basic or QBasic, with a REPL as well would do quite nicely.

On the other hand maybe what is missing is making those little boards with JavaScript or Python more widespread.