The current implementation in GRPlatformTest attempts to guess whether the Smalltalk dialect supports Unicode, but it raises an exception in platforms such as VAST where it does support it but not by using Characters with codepoints greater than 255 (e.g. VAST uses a Grapheme class for that).

Te guesswork could remain, but it should be moved to the platform specific class, so we don't have to patch that code in the original codebase and we can determine whether we support it in our own packages.

supportsUnicode
	"dynamically try to figure out whether the current dialect supports Unicode"
	^ [ 
		String
			with: (Character value: 16r1F1F3)
			with: (Character value: 16r1F1F1).
		true
	 ] on: Error do: [ :error | false ]