Chapter 8 of 9
Conclusion
The evidence separates what Perl 6 got right from what it transmitted. Its specification picked several answers the industry later converged on, five to twenty years early — the closing-delimiter heredoc rule and defined-or in 2001, success/failure scope blocks, /x by default and grammars as classes in 2002, named-only parameters in 2003 — but almost none of that judgment turned into citations, because ideas travel with shippable implementations and with the languages designers already use. The heredoc rule, Perl 6's 2001 choice, is the one design decision that shipped intact in a major non-Perl language, and even there the credit survives only in the swift-evolution thread, not in SE-0168. What did cross the boundary, in five documented borrowings and Perl 5's many, was small, separable syntax or semantics that a host language could adopt without buying into the rest: an arrow token, a heredoc margin rule, a string-increment rule, a regex whitespace default, a CPAN object system, and below the language level a lookbehind technique and a separator modifier. The features that demand whole-language commitment — junction autothreading, hyper and meta-operators, Whatever-currying, NFG, grammar subclassing, slangs — have not moved at all, and their non-adoption measures how much of Raku is architecture rather than syntax.
Two practical lessons follow for a Raku implementer. Priority claims need the right date: the Apocalypse establishes the idea, but Pugs (2005) and Rakudo (2009–2015) decide whether anyone outside could have used it, and late-specified features such as :D, NFG and Supplies were contemporaries of their analogues, not ancestors; even sub MAIN, specified four years before plac, became usable in full only the month plac appeared. And Perl 5's smartmatch is the case study in borrowing too early — it froze a 2002–2007 snapshot of a design Raku then revised, drew Larry Wall's own apology within weeks of shipping, and spent eighteen years unable to fix or remove it. The pieces of Raku now being cited in PHP, Ruby, Swift, C#, Carbon and TC39 discussions — range endpoints, rational defaults, chained comparison, regex ergonomics — are exactly the settled, separable ones, which suggests that stable and well-documented semantics for such pieces, rather than the ambition of the whole design, are what other languages will continue to pick up.