Re: continuations in Scheme

[Axel Kittenberger <axkibe@...>]

> think like coroutines that aren't tied to a function but to the whole
> execution. and that can not only return back and forth, but also
> branch... it's a quite powerful primitive

Thats why I reckoned it would be very intune with the Lua design
goals: generic and powerful [1]* And if it could made to also be able
to do a (named) next loop iteration in a concise way, that way I would
get what I (and others) want, Robertos design goals fullfilled and
everybody would be happy. I'd not speaked about it unless I get
something running one day, but this might take long and is too complex
for me, but the GSoC seemed like a good place to inject this. Have to
ask the computer scientists, but might that be worth a paper as well?
While call-cc is an old hat due to Scheme, I don't know of procedual
languages embracing it. I didn't do any re/search about this claim
tough.

[1] http://lua-users.org/lists/lua-l/2011-01/msg01688.html
* Continuations are also more powerful than gotos anyway. And in
footprint size I suppose it would be roughly equal sized to
coroutines, which would become more a special case of continuations
than a need of support by themselves.