On Mar 20, 2008, at 8:54 AM, Lars Hansen wrote: > We have set aside placeholders for E4X syntax. How useful this is I > don't know; the experience with reserving future reserved words in ES3 > has been mostly negative (as a rule programmers don't read specs, and > when they do, as a rule they ignore the "may be used in the future" > clauses -- and I don't think they're wrong in doing so).
Generally I agree. Just for the record, the "future reserved words" go back to ES1 and were prefigured by Netscape's reserving all then- reserved Java identifiers. This was done with agreement and a great deal of spec-writing leadership by Microsoft, but the JScript engine nevertheless reserved only class, enum, extends, and super (if memory serves). Over time, as Netscape went into decline, content grew to use identifiers such as 'char'. And of course we are contextually unreserving in ES4 (Firefox 2 / JS1.7 already does this), but it wouldn't help the 'char' case I recall, where the identifier was a parameter name. This tale cautions about several things other than trying to reserve future syntax, among them the participants in the standard not following through early in their own products, before conflicts in the market could emerge. /be _______________________________________________ Es4-discuss mailing list [email protected] https://mail.mozilla.org/listinfo/es4-discuss
