For a long time we’ve said it’s important to separate semantics and presentation (eg html vs css).
I’ve always wondered why we never followed our own advice when creating new programming languages. Let the IDE present whatever TF you want, and in the background it’s simply building an Abstract Syntax Tree and serializing that to a file in whatever serialization format it prefers.
I’ve dreamed of making my own data-flow language that follows that principle, but I have neither the knowledge of compilers nor the free time to try.
For a long time we’ve said it’s important to separate semantics and presentation (eg html vs css).
I’ve always wondered why we never followed our own advice when creating new programming languages. Let the IDE present whatever TF you want, and in the background it’s simply building an Abstract Syntax Tree and serializing that to a file in whatever serialization format it prefers.
I’ve dreamed of making my own data-flow language that follows that principle, but I have neither the knowledge of compilers nor the free time to try.