Re: issue IN-PACKAGE-FUNCTIONALITY

[masinter.pa@Xerox.COM]

My belief is that we can fix the "problem" stated in this issue--namely
that the current dual purpose of IN-PACKAGE reduces error checking--by
submitting a proposal as follows:

a) this proposal should say only:

  Eliminate the ability of IN-PACKAGE to create a package on demand.
  Eliminate the :NICKNAMES and :USE arguments to IN-PACKAGE, since they
  are no longer needed.


b) the DEFPACKAGE issue itself should contain the phrase 
"Clarify that DEFPACKAGE is the preferred way to declare a package,
  and MAKE-PACKAGE is the preferred way to construct a package at runtime."

c) The proposal

 "Require IN-PACKAGE to signal an error if the package does not exist."

should be changed; instead, the results are unspecified. Note that
implementations of CLtL might indeed create the package. This is exactly
because of the compatibility issue.

d) the proposal
"Eliminate the compile-time processing requirement for all package-related
  functions except IN-PACKAGE and DEFPACKAGE."

should be withdrawn at this time to be dealt with more uniformly by the
compiler committee.