Problem: There is no conventional way for applications to identify the version of the parser they are using, for purposes of diagnostics or other kinds of troubleshooting.
The best the JVM supports is the JDK 1.2
which is dependent on the JAR file metadata. It provides a partial solution, at
the price of portability (JDK 1.1 APIs are much more portable) and
assumptions like "one parser per package".
Solution: Define a new standard read-only property:
Returns a string identifying the reader and its version for use in diagnostics.
Parsers could support that if desired, probably using some sort of resource-based mechanism (not necessarily Package) to keep such release-specific strings out of the source code.
Open issue: Should there be separate strings to ID the reader (likely a constant value) and its version (ideally assigned in release engineering)?