I found a minor grammar error in the posting yesterday; patch to fix is below (which I'll put in a later update).

The last version includes some interesting citations related to precedence, in particular a quantitative study.  I think they give strong evidence for the reasonableness of the approach, so I pulled them in.

I don't know of anything else that needs doing; my intent is for the next revision to be the last one.  Comments?

--- David A. Wheeler

--- a/SRFI-105.html
+++ b/SRFI-105.html
@@ -1002,8 +1002,8 @@ higher precedence than division.
 Dennis M. Ritchie, in his
 <a href="http://cm.bell-labs.com/cm/cs/who/dmr/chist.html">
 <i>The Development of the C language</i></a>,
-notes that of the precedence rules were infelicitous;
-should they be used or not?
+notes that some of the precedence rules were infelicitous;
+should those problematic rules be used or not?
 There would also be substantial disagreement on exactly what operators
 should be in the precedence table (including which combinations and if
 Unicode characters should be included), their order, and whether