[idn] stringprep comment 4 : hangul jamo filler

["Soobok Lee" <lsb@postel.co.kr>]

hangul conjoining jamo filles look like blank space when they don't form  complete 
hangul syllables. 

I had suggested in my hangulchar draft that such standalone fillers should be prohibitted
by new context-sensitive procedural prohibition stage after table-mapping based
current prohibition stage. (In the same position like BIDI checking stage suggested by Martin
in the previous posting)

The very suggestion was abandoned with all other bypassing suggestions for 
other UAX15 shortbacks. 

Fillers may be the security holes of IDNA if not taken carefully in loosely managed zones.

Soobok Lee