To me, dependent types are more "intuitive" than objects, since objects are essentially arbitrary, whereas types are central to the sense of the semantics. I think I'll give this language a try.
I'm tempted to design a compact nonfunctional programming language and call it Pure Idiocy. It would be a pre-processor for any other language. 😉. Even after final compilation in any other language, a program would always be recognized immediately as having its roots in Pure Idiocy.
I know my ideas may sound idiotic, but I contend that they are merely naive. And, one of the virtues of naivete is that I do not have an already deeply ingrained way of thinking.
That said, I've been quite disappointed with "object orientation". I was a very early Python adopter - and people made fun of me when I tried to explain how the indentation worked.... and later, when it became more and more object-oriented I came to realize that it was not easier, and I was not more productive, and that classes and objects are not very powerful means of abstraction - and to me are more like an additional layer on top than an means of simplification. Correct me if I am wrong, but objects are still operating essentially at the level of first-order logic.
So, if one is to approach abstractions of data and program (together), one should do this in terms of functions and types, and then one may gain some ease of access to higher-order approaches to problem-solving.
tom grzyb
in reply to tom grzyb • • •Brian Fitzgerald
in reply to tom grzyb • • •tom grzyb
in reply to tom grzyb • • •tom grzyb
in reply to tom grzyb • • •I know my ideas may sound idiotic, but I contend that they are merely naive. And, one of the virtues of naivete is that I do not have an already deeply ingrained way of thinking.
That said, I've been quite disappointed with "object orientation". I was a very early Python adopter - and people made fun of me when I tried to explain how the indentation worked.... and later, when it became more and more object-oriented I came to realize that it was not easier, and I was not more productive, and that classes and objects are not very powerful means of abstraction - and to me are more like an additional layer on top than an means of simplification. Correct me if I am wrong, but objects are still operating essentially at the level of first-order logic.
So, if one is to approach abstractions of data and program (together), one should do this in terms of functions and types, and then one may gain some ease of access to higher-order approaches to problem-solving.
tom grzyb
in reply to tom grzyb • • •