(A correspondent writes:)

I may have found a bug in JVMS 11, section 4.7.9.1.

Page 130 of http://cr.openjdk.java.net/~iris/se/11/latestSpec/java-se-11-jvms-draft-diffs.pdf says:

> A class type signature represents a (possibly parameterized) class or interface
> type. A class type signature must be formulated such that it can be reliably
> mapped to the binary name of the class it denotes by erasing any type arguments
> and converting each . character to a $ character.

An example of a class type signature is
  Lfoo/bar/Baz<Lp1.Quux;>.22;
Doing the mapping described in JVMS yields
  Lfoo/bar/Baz$22;
which is a field descriptor (JVMS §4.3.2), not a binary name (JLS §13.1).
The corresponding binary name would be
  foo.bar.Baz$22
and the internal form of the binary name would be
  foo/bar/Baz$22

PS: Also a suggestion: on page 354 (twice), page 357 (once), and page 365 (once), change text like "the name is the binary name" to "the name is *the internal form of* the binary name", for clarity.