lambda, heap and stack

[Julian Rohrhuber]

Hi, one very basic question to which I could only find unclear answers on the net so far. I’m pretty sure you can help me.

My original understanding was that in order to implement fully first-class functions with lexical scope, you need to go beyond the simple stack memory architecture. The reason is that if you return a function from a function, this will be called outside the nesting structure of the stack. But it seems that somehow C++11 does create lambdas on the stack? How is it done in xtlang?

Best,
Julian

[Andrew Sorensen]

Hey Julian,

Short answer, memory zones.  Also known as regions or arenas.  I made a short high level video if you're interested.

https://www.google.com.au/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=0ahUKEwj-0Im8qKPWAhXDopQKHfXzBbwQtwIIKDAA&url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DkxD7-Pf_ZyY&usg=AFQjCNHxVLlFWS5-2liFQ97oxaHvCBh2bg

Cheers,

Andrew.

--
You received this message because you are subscribed to the Google Groups "Extempore" group.
To unsubscribe from this group and stop receiving emails from it, send an email to extemporelang+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Julian Rohrhuber]

Hey Andrew,

very good, I get it, up to a point. Your video explains the basics very well.

Once we ask for returning closures from closures, is it right that it is not exactly trivial anymore?

I read a little in Tofte and Talpin 1994, and Aiken et. al 1995, and it seems to be quite a bit of reasoning that must go on under the hood!

Cheers,
Julian

> To unsubscribe from this group and stop receiving emails from it, send an email to extemporelan...@googlegroups.com.

> For more options, visit https://groups.google.com/d/optout.
>
>

> --
> You received this message because you are subscribed to the Google Groups "Extempore" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to extemporelan...@googlegroups.com.

[digego]

Hey Julian,

No, not trivial, at least from a 'human' reasoning perspective.  

[Julian Rohrhuber]

> On 22.09.2017, at 03:36, digego <dig...@gmail.com> wrote:
>
> Hey Julian,
>
> No, not trivial, at least from a 'human' reasoning perspective.

So that means that in xtlang we do have a memory structure for this that goes beyond a simple stack?

Thanks!