Hello,

In preparation for the next meeting, I spent some time doing some work
on a new revision of this paper. Last time we discussed it we had a
bit of a hard time with the overall framing of the problem the paper
is trying to solve, so this new revision introduces a new section that
goes through a bit of the background requirements, which hopefully
should help guide the conversation better.

Changes are:

 * Change “Binary Module Interface” to “Built Module Interface”.
 * Introduce a new section exploring the outcome of having independent
parsing context between the translation unit importing a module and
the translation unit declaring a module interface.
 * Editorial changes throughout the paper for better readability,
including removing redundancy with the new section.

Revised version at https://wg21.link/p2581r1

daniel