2 ms·
> The documentation size is also quite low You can guess that it is very declarative code, given that they used a Lisp with high-level facilities on top.
by lispm 10y ago
> The documentation size is also quite low
You can guess that it is very declarative code, given that they used a Lisp with high-level facilities on top.
- marcosdumay 10y agoI'm not going back into the article, but there's a "code reads like documentation" dimension on the evaluation. I don't remember it getting a top score there.
- lispm 10y agoBut got an A in the table for "Executable Specification", that's what 'declarative programming' actually means. One specifies/declares WHAT to do and the system figures out HOW to do it. Relational Lisp supports this on a 'logical'/'relational' level.
- kazinator 10y agoAnything declarative is pretty much going to be unreadable to the average coder, who equates readability with "able to follow the Fortran-like step-by-step recipe of what the computer is being told to do or calculate". If you've had a 30 year career writing imperative, OOP and functional code, with zero declarative experience (logic programming and that sort of thing), it's not magically going to be readable to you. You have to, ... ahem! ... backtrack a bit, and go down a different learning path. It's somewhat telling that the Lisp shared an A score in the "Executable Specification" area with Haskell and that Proteus. The subjective panel could tell that there is a wonderful executable specification there; they just didn't quite grok its notation or semantics. :) Or maybe they did grok it: they started with an A+, and then docked points down to a B for all the meddlesome parentheses. :)