Amend RFC1228 with operator fixity

Author: nagisa

text/1228-placement-left-arrow.md

@@ -168,16 +168,22 @@ lowest precedence) to highest in the language. The most prominent choices are:
 1. Less than assignment, as proposed by this RFC:
 
     Assuming `()` never becomes a `Placer`, this resolves a pretty common
-    complaint that a statement such as `let x = y <- z` is not clear or readable
-    by forcing the programmer to write `let x = (y <- z)` to typecheck.
+    complaint that a statement such as `x = y <- z` is not clear or readable
+    by forcing the programmer to write `x = (y <- z)` for code to typecheck.
 
-2. More than assignment and binop-assignment, but less than any other operator:
+2. Same as assignment and binop-assignment:
+
+    `x = y <- z = a <- b = c = d <- e <- f` parses as
+    `x = (y <- (z = (a <- (b = (c = (d <- (e <- f)))))))`. This is so far
+    the easiest option to implement in the compiler.
+
+3. More than assignment and binop-assignment, but less than any other operator:
 
     This allows for various expressions involving assignment to be parsed
-    reasonably. For example `let x = y <- z += a <- b <- c` would get parsed as
-    `let x = ((y <- z) += (a <- (b <- c)))`.
+    reasonably. For example `x = y <- z += a <- b <- c` would get parsed as
+    `x = ((y <- z) += (a <- (b <- c)))`.
 
-3. More than any operator:
+4. More than any operator:
 
     This is not a terribly interesting one, but still an option. Works well if
     we want to force people enclose both sides of the operator into parentheses