4 ms·

I see both of those things a lot. The problem is if you aren't already familiar with those idea, the statements mean literally nothing. They basically don't mak

I see both of those things a lot. The problem is if you aren't already familiar with those idea, the statements mean literally nothing. They basically don't make sense.

ken 7y ago
These examples aren't so far out there. DSLs exist in other languages. Statements exist in other languages. A programmer who has never used macros can imagine being able to create these, even if they have no idea what exactly that would look like in a Lisp program. That issue is also true of any game-changing technology you haven't used yet. If you hadn't used a personal computer, we could enumerate the features and advantages, but you probably wouldn't be able to truly appreciate it until you used one yourself.
AlexCoventry 7y ago
They're also rarely useful things to do in regular software development.
- quickthrower2 7y ago
  I wonder if we struggle needlessly though in "regular software development" - and entire jobs or teams exist because we didn't find the right way to express things in a language. OTOH you could create lisp code that is so unique and weird in it's semantics that it's hard for a new team member to get onboard. But I've seen that done in C# too which allows you to have code that writes code at runtime (not as elegantly though!).
  - AlexCoventry 7y ago
    Yeah, writing your own language and self-modifying code can give you a lot of leverage, but usually they also make the code so much harder for other people to pick up that it's not worth it.
FrozenVoid 7y ago
The idea is that all LISP code is data that is easily altered, essentially https://en.wikipedia.org/wiki/Abstract_syntax_tree https://en.wikipedia.org/wiki/Abstract_syntax_tree encapsulated in parens. LISP syntax is uniform,everything(functions,data,argument lists) is nested lists of lists like (list(list(list))), so that allows altering specific places "in-the-list"(place 1,2,3...) as form of meta-programming, where normal syntax FOR( A IN B) would be written roughly as (for a b) where for is used as a function taking 'a' and 'b' as parameters, but in this form(where 'for' is just the first element of the list) you can have functions that "alter-the-list" by example (replace-first-list-element (for a b) replacement-part) returns (replacement-part a b).

I see both of those things a lot. The problem is if you aren't already familiar with those idea, the statements mean literally nothing. They basically don't make sense.

ken 7y ago

These examples aren't so far out there. DSLs exist in other languages. Statements exist in other languages. A programmer who has never used macros can imagine being able to create these, even if they have no idea what exactly that would look like in a Lisp program. That issue is also true of any game-changing technology you haven't used yet. If you hadn't used a personal computer, we could enumerate the features and advantages, but you probably wouldn't be able to truly appreciate it until you used one yourself.

AlexCoventry 7y ago

They're also rarely useful things to do in regular software development.

quickthrower2 7y ago
I wonder if we struggle needlessly though in "regular software development" - and entire jobs or teams exist because we didn't find the right way to express things in a language. OTOH you could create lisp code that is so unique and weird in it's semantics that it's hard for a new team member to get onboard. But I've seen that done in C# too which allows you to have code that writes code at runtime (not as elegantly though!).
- AlexCoventry 7y ago
  Yeah, writing your own language and self-modifying code can give you a lot of leverage, but usually they also make the code so much harder for other people to pick up that it's not worth it.

FrozenVoid 7y ago

The idea is that all LISP code is data that is easily altered, essentially https://en.wikipedia.org/wiki/Abstract_syntax_tree https://en.wikipedia.org/wiki/Abstract_syntax_tree encapsulated in parens. LISP syntax is uniform,everything(functions,data,argument lists) is nested lists of lists like (list(list(list))), so that allows altering specific places "in-the-list"(place 1,2,3...) as form of meta-programming, where normal syntax FOR( A IN B) would be written roughly as (for a b) where for is used as a function taking 'a' and 'b' as parameters, but in this form(where 'for' is just the first element of the list) you can have functions that "alter-the-list" by example (replace-first-list-element (for a b) replacement-part) returns (replacement-part a b).