Lars T. Hansen wrote:

<his own summary of how he saw the discussion>

Here's how I summarize the discussion: SRFI-15 is a summary of an
often-used 20-line macro hack to throw together something that looks like
dynamic variables that (a) doesn't work in multi-threaded environments,
and (b) requires all dynamic variables to have global lexical scope.
When people point out these issues, Hansen replies, basically, that this
is what he intended, that he just wants to fix a standard implementation
technique, and that he gets his semantics from whatever this standard
implementation technique gives him.

That's OK, but I don't see what progress this SRFI makes.

Brad Lucier