as the old code. And once that works well, we can remove the old code
and old fields.
+** Implement Unicode-compliant display of "default-ignorable" characters
+See the "Characters Ignored for Display" section of paragraph 5.21 in
+the Unicode Standard for the details.
+
+The implementation would import the data from Unicode UCD file
+DerivedCoreProperties.txt, and provide a minor mode that arranges for
+the characters with the Default_Ignorable_Code_Point (DI) property to
+be hidden on display. One way of implementing that could be via
+glyphless-char-display-control; that one is global, but maybe there's
+a way of making it buffer-local. Alternatively, this could be
+implemented in C in the display engine.
+
+An additional aspect of this is the display of U+00AD SOFT HYPHEN as
+invisible except at line boundaries. Note that this would need to
+support hard (physical) newlines in the buffer as well as soft
+wrapping of long lines under 'visual-line-mode'. The algorithm for
+selecting the wrap point may also need be changed to break at the soft
+hyphen.
+
** FFI (foreign function interface)
See eg https://lists.gnu.org/r/emacs-devel/2013-10/msg00246.html