On Jul 19, 2011, at 2:44 PM, Allen Wirfs-Brock wrote:

> These were largely topics of another thread which I want to get back to with 
> some new thoughts.  But these issue may have impact of details of the direct 
> of the design discussion in this thread.  I don't think it works to try to 
> add private property access issues into the design after everything else is 
> settled.

(Missing "until" before "after"?)

Good point, private should not be "syntax solved" first. We need to finish the 
design, including semantics. Basing class-private instance and even prototype 
properties on private name objects still makes more sense to me than any new, 
unobservably different way of specifying private-in-class. Indeed 
private-on-class-prototype may leak some observable difference (but not the 
private name object) via Proxies, I'm still not sure.

Another reason to separate private from classes, as dherman's "minimal classes" 
post suggested.


es-discuss mailing list

Reply via email to