Re: Perl Implementation Language

[Nathan Torkington]

Ken Fox writes:
> The dogfood theory? One of the design goals for Perl is to make text
> munging easy. Parsing falls into that category and therefore we should
> use it, i.e. eat our own dogfood.

How about this.  During design, we try to make the parser a module
with an interface designed so that it could be done via Perl
callbacks.  At this stage we should be able to say "whoa!  This is
crazy, we can't do it", or "yes, we can leave the door open for
this."

I think we are going to have enough on our plates without trying to
use a language that doesn't exist to write the parser for that
language :-)

That kind of bootstrapping problem adds a lot of uncertainty to the
development, and another level of implementation to overcome.  I'd
be happy to make it *possible* to write the parser in Perl and then
start doing that in 6.1.0.

Nat