Dynamic Language Runtime

David Bovill david at openpartnership.net
Wed May 9 11:21:09 EDT 2007


Thanks Mark. In general you seem to be supporting the potential of this
path? It seems at this stage more promising than other stabs at this no -
say Open Scripting Architecture (OSA), and yes it means a lot more than just
a browser plugin - though i guess this would get most peoples attention
here?

A question about "writing a transcript compiler for DLR"? What does this
involve - starting from scratch in C# or if you were RunRev and already
presumably had some code basis for a compiler would you be able to use that
directly - I guess not. Secondly if it means starting from scratch would
that mean anyone could legally undertake such a thing - that is to write a
minimal Transcript compiler for the DLR. Its not something I am picturing
doing - but I am curious :)

On 09/05/07, Mark Wieder <mwieder at ahsoftware.net> wrote:
>
> David-
>
> I was part of a DLR roundtable discussion at CommunityOne day at
> JavaOne yesterday, so I think I can speak to this from the bleeding
> edge. A few points to consider:
>
> 1. The DLR is alpha. I'm bullish on it, and it's very promising, but
> there are quite a few rough edges and things that just aren't worked
> out yet. There's no consensus on threads, for instance. Give it time.
>
> 2. The DLR means a lot more than just being able to run something in a
> browser. It promises cross-language compatibility, the ability to run
> modules written in one language within the context of another
> language.
>
> 3. There's never been a jit compiler for xtalk. Not that it couldn't
> be done, but that brings headaches of its own. The whole "short path /
> long path" thing, for one, is something that I don't think has ever
> been considered for xtalk syntax.
>
> 4. Peter Fisk's Smalltalk compiler is "capable of processing
> arithmetic expressions". That's still a far cry from a full-blown
> Smalltalk compiler. I'm reasonably certain that a compiler that would
> process "put 2 + 3" in Transcript wouldn't be that hard to build. But
> there's a lot more to the infrastructure than that.
>
> 5. We can't all be Peter Fisk.



More information about the use-livecode mailing list