[Cfp-interest] signaling NaN I/O issue?

[David Hough CFP]

> Printf can't return a character string representation 
of a signaling NaN, because there isn't one.

Since we considered the possiblity of making signaling NaN conformance
a separate option from the rest of IEEE conformance,
have we considered this possibility:

an implementation supporting signaling NaNs shall read and write,
without signaling an exception,
an implementation-defined signaling NaN character sequence,
rather than signaling and quieting?

or that could be a "should" instead of "shall", and then 
an implementation supporting signaling NaNs shall signal and quiet if
it won't read and write?