* whether E's features will (could) be supported in Perl 6
* whether E could be supported by Parrot
Here are some references, for the curious:
E language home page
http://www.erights.org
The E Language in a Walnut
http://www.skyhunter.com/marcs/ewalnut.html
-r
--
email: r...@cfcl.com; phone: +1 650-873-7841
http://www.cfcl.com - Canta Forda Computer Laboratory
http://www.cfcl.com/Meta - The FreeBSD Browser, Meta Project, etc.
http://www.ptf.com/dossier - Prime Time Freeware's DOSSIER series
Syntacticaly speaking, it would certainly be easy to write E-ish
looking code in Perl 6. It would be a little harder to limit some
of the semantics that E puts limits on (such as no mutable statics),
but undoubtedly it's doable with some amount of effort. As a limiting
case, you can always back out the entire Perl grammar and install
the E grammar in its place...
My job as a language designer is to tell people what they can do,
not what they can't do. But one of the things people can do is to
tell their own selves what they can't do. I think "use Eish" falls
into that category.
: * whether E could be supported by Parrot
Well, obviously, E is supported by the JVM currently, so the answer is
probably "yes" at some level. The "promises" model is just message
passing in disguise, so that part is probably fairly trivial--to
the extent that doing anything with encrypted links can be called
"trivial", which is not very. Certainly Parrot is being designed
to handle all the asynchronicity properly. I suppose the main
question is to what extent Parrot supports the capability model of
security natively. With E on the JVM, it seems to be rather bolted
onto the side. Minimally, the same could be done for Parrot. But my
smudgy memory tells me that Dan was going for more of a capability
based model natively for Parrot. I could be confabulating though.
Anyway, thanks for the interesting link. (But if you want an opinion
from the Parrot folks, you'll probably have to ask again over on p6i.)
Larry