Continuations are not so good, delimited continuation are far better

Continuations are not so good, delimited continuation are far better

[Yannick Duchêne (Hibou57)]

Hi out there,

I'm posting this here, after some previous talks about the wish to get  
coroutines in Ada.

Coroutines are related to continuations. Continuation is well know  
abstraction in the functional programming world, which often forget about  
real world limitations, something Ada is necessarily aware of. While  
mathematically correct, continuations are not so good in practice.

I was questioning myself about continuation's implementation in SML  
compilers (like MLton), and landed onto something worth to be forwarded in  
an Ada newsgroup: a multi‑page document about continuations, where the  
author argues against classical continuation and argues in favour of  
delimited continuation. Somewhere in these pages, he notes an  
implementation or compiler, will anyway have to determine a delimited  
continuation for each continuation, to implement continuations  
efficiently. That's what I had in mind too, and that's why I enjoy this  
document (I have not read all of it yet).

See: [Continuations and delimited  
control](http://okmij.org/ftp/continuations/index.html)

-- 
“Syntactic sugar causes cancer of the semi-colons.” [1]
“Structured Programming supports the law of the excluded muddle.” [1]
[1]: Epigrams on Programming — Alan J. — P. Yale University