On Thu, 2013-01-03 at 18:39 -0500, John Cowan wrote:
> Am I right to say that given the
> R6RS (as opposed to Racket) rule that an identifier cannot be defined
> differently at different phases, that the correct implicit phase can
> always be deduced, and that therefore the machinery of explicit
> phasing
> is not actually required?  Or do you think that's a fallacy?
> (Or am I misstating what R6RS allows?)

I think I would have to go back to Aziz's dissertation on that one to
know for sure. My inclination is to say that yes, this is correct. In
particular, if you allow an identifier to reference only a single
binding over the course of its entire lifetime across all phases, then
either the identifier is needed at a given phase, or it is not, and this
can be determined implicitly. The reason that this does not work if you
allow for multiple bindings to the same identifier for each phase is
that you must then somehow explicitly identify which binding goes with
which phase, and this cannot be deduced by the implicit phasing system.

--
Aaron W. Hsu | arcfide@sacrideo.us | http://www.sacrideo.us
לֵ֤ב חֲכָמִים֙ בְּבֵ֣ית אֵ֔בֶל וְלֵ֥ב כְּסִילִ֖ים בְּבֵ֥ית שִׂמְחָֽה׃

_______________________________________________
Scheme-reports mailing list
Scheme-reports@scheme-reports.org
http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports