An "extend" operator is a natural companion to <|

Brendan Eich brendan at
Tue Jul 19 13:35:58 PDT 2011

On Jul 19, 2011, at 1:28 PM, Luke Hoban wrote:

> Agreed - my note was just a "most value to the most users the soonest" argument.  Syntax is clearly important, and many of the proposals can only reasonably be offered as syntax.  Some of the proposals that can be partly delivered as libraries still benefit significantly from syntax as well.

Great to hear.

>  But the adoption timelines for the new library pieces and the new syntax pieces will be quite different, not just because of downrev browsers, but also because of breaking change burden, authoring convenience, maintenance of old script, etc.

Not to quibble, but the problem you are identifying is the polyfill vs. compiler one: polyfills for libraries are easier (even if fake or weak, semantically), polyfills for syntax are transpilers or full-strength compilers. Right?

I'm hoping your point is not that some implementors will prototype library code faster than new syntax. Library extensions may be easier (proxies are a possible counter-example) but implementors should be prototyping both new syntax+semantics and new library code or built-in modules (which will require the module system).


More information about the es-discuss mailing list