Re: Not revisiting the RFC process (was: RFC 362...)

[Nathan Torkington]

What bothers me about the idea of reopening the RFC proces is that the
RFCs we have are already overwhelming.  I think one reason we haven't
seen anything from Larry yet is that the RFCs cover an enormous area,
but in a disconnected and difficult to summarize and use form.

I don't think there's any harm in suspending the RFC process for a
while.  Nobody denies that some modified RFC process will be needed as
we go on, to codify and archive discussion, but I think there are good
reasons for not reinstituting it just yet.

First, Larry's overwhelmed by the current set.  Adding more won't make
this any better.

Second, the current RFC process had flaws.  Many people ragged on it,
and I'm happy to admit it wasn't perfect.

Third, good ideas aren't going to go away just because we focus our
energies on building perl6.  They'll still be there when we have
perl6, and will be more easy to deal with because then we'll actually
have something to measure them against.

So I think energy would be best spent on helping Dan with the design.
If you don't feel you're up to that, then by all means work on the RFC
process.  But I think the RFC process needs rethinking, rather than
reopening.

And I don't want to do that rethink yet.  It will be a huge affair,
with sixty people all offering their great ideas and arguing with one
another, and ultimately that's all secondary to getting an alpha of
perl6, which is my main goal.

So I'd like you to sit on your good ideas.  Write them down, by all
means, just don't expect us to be able to doing anything with them
yet.  My energies are all focussed on finding a way to get a language
spec out of Larry.  Anything outside of that is a distraction.

In fact, it occurs to me that once Larry starts emitting ideas on
perl6, you'll be able to subtly introduce your thoughts.  Nobody will
spank you because they weren't in the RFCs.  The point of the RFCs was
to get a body of thinking that roughly defines the areas of concern
for perl6.  To some extent it will have elements of changing the spec
after people have begun work, but I don't think anyone has ideas for
Perl that would be outside the parameters of what Larry's thinking :-)

Nat