Another NetRexx wannabe?

[Fernando Cassia-2]

Julien Ponge, who, in addition to being a Java developer and a
professor, also writes technical articles for both otn/java and Java
Magazine, has created Golo, a simple, dynamic, weakly-typed open
source language that favors the explicit over the implicit. Developers
can pick it up in a manner of hours, not days. Responses to its recent
release at Devoxx have been favorable.

Built from day 1 with invokedynamic, and currently in beta, Golo takes
advantage of the latest advances of the JVM. It is also a showcase on
how to build a language runtime with invokedynamic.

https://blogs.oracle.com/java/entry/golo_a_lightweight_dynamic_language

I'm interested in knowing what "built from the ground up with
inokedynamic" (introduced in JDK7) compares with NetRexx approach
(which AFAIK developed with JDK 1.2+ in mind).

Would it be possible, and how difficult, to build a "NetRexx with
invokedynamic" in mind?

Perhaps MFC or Rene could comment?.
Thanks,
FC

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

[ThSITC]

Hi Fernando,

1.) Thanks for this X-info!
2.) My (personal) opinion, is, that *Rony Flatscher's* BSF4ooRexx, when
ported from ooRexx
to NetRexx, shall already *now* contain *any and all of the*
invokedynamics feature You do
quote in your mail below!

Rony, Rene, all:

*When I'm wrong*, please do say so ...

Happy sunday, anyway!
Thomas.
====================================================================
Am 07.04.2013 03:12, schrieb Fernando Cassia:

> Julien Ponge, who, in addition to being a Java developer and a
> professor, also writes technical articles for both otn/java and Java
> Magazine, has created Golo, a simple, dynamic, weakly-typed open
> source language that favors the explicit over the implicit. Developers
> can pick it up in a manner of hours, not days. Responses to its recent
> release at Devoxx have been favorable.
>
> Built from day 1 with invokedynamic, and currently in beta, Golo takes
> advantage of the latest advances of the JVM. It is also a showcase on
> how to build a language runtime with invokedynamic.
>
> https://blogs.oracle.com/java/entry/golo_a_lightweight_dynamic_language
>
> I'm interested in knowing what "built from the ground up with
> inokedynamic" (introduced in JDK7) compares with NetRexx approach
> (which AFAIK developed with JDK 1.2+ in mind).
>
> Would it be possible, and how difficult, to build a "NetRexx with
> invokedynamic" in mind?
>
> Perhaps MFC or Rene could comment?.
> Thanks,
> FC
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
> Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
>
>

--
Thomas Schneider, IT Consulting; http://www.thsitc.com; Vienna, Austria,
Europe

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

[rvjansen]

Hi Fernando,

invokedynamic has taken its time to arrive. At the 2008 JVM symposium (where I was still thinking that open source NetRexx would be a matter of weeks) all the talk was around invokedynamic; indeed its inspiration was to take the overhead out of dynamic languages on the JVM. Dynamic used here in the sense of the user being able to define methods on the fly, and have a more sensible way of type checking, like demonstrated with the Visitor (anti-)pattern.

I have followed invokedynamic closely for some time, but it became a moving target and a Java-level api has taken a long time; to use it, for a long time one had to assemble bytecode instructions. This, and the fact that NetRexx uses this approach in only a ver small part of the interpreter, has thrown me off, and to give you a more valuable answer, I  would have to delve into it anew.

Suffice it to say, that my current view is, that invokedynamic is a solution for a problem that NetRexx currently does not have, i.c. bad performance on dynamically constructed calls. The fact that one has to use the interpreter for this, and from-memory compilation was only added recently (3.01) makes for very few use-cases that can be studied.

Thanks for drawing our attention to this, and I will certainly take a look at it.

best regards,

René.
On 7 apr. 2013, at 03:12, Fernando Cassia <[hidden email]> wrote:

> Julien Ponge, who, in addition to being a Java developer and a
> professor, also writes technical articles for both otn/java and Java
> Magazine, has created Golo, a simple, dynamic, weakly-typed open
> source language that favors the explicit over the implicit. Developers
> can pick it up in a manner of hours, not days. Responses to its recent
> release at Devoxx have been favorable.
>
> Built from day 1 with invokedynamic, and currently in beta, Golo takes
> advantage of the latest advances of the JVM. It is also a showcase on
> how to build a language runtime with invokedynamic.
>
> https://blogs.oracle.com/java/entry/golo_a_lightweight_dynamic_language
>
> I'm interested in knowing what "built from the ground up with
> inokedynamic" (introduced in JDK7) compares with NetRexx approach
> (which AFAIK developed with JDK 1.2+ in mind).
>
> Would it be possible, and how difficult, to build a "NetRexx with
> invokedynamic" in mind?
>
> Perhaps MFC or Rene could comment?.
> Thanks,
> FC
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
> Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
>

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/