parrot-dev Digest, Vol 58, Issue 8

[Bart Wiegmans]

Hi everybody,

Let us not get angry at one another, for one thing. I personally think
the question of the future direction of parrot, if any, is a vital
one. Because no one can deny that with the advent of rakudo/jvm and
MoarVM, the dynamic has changed. Why? Because rakudo was our biggest
'customer', so to speak. Parrot and Perl6 started out together, and
now that they have finally truly separated, it is only natural to
reconsider why we are still around.

And that is not such a simple question to answer. Parrot is supposed
to be the VM for all dynamic languages. By any reasonable definition,
this goal has not been achieved. In fact the opposite: implementations
of dynamic languages have proliferated over the last few years. The
main problems of dynamic languages in the past (slow, no real thread
support, huge duplication of effort) are alive and well today. (Except
perhaps slowness, in a few select instances, but that is another
matter).

Parrot can / could / should have solve(d) these problems, but has not.
Why not? I don't pretend to know. Is it still even important, now that
the trends have moved away from dynamic to static languages (go,
scala) and server to mobile / browser? Again, I don't know.

What I do believe, though, is that the original goal for parrot is
still worthy, and is still unmet. Parrot should be itself, but - of
course - much better :-). Dynamic languages shouldn't suck. Go team!

Regards,
Bart Wiegmans