What you're asking for, then, is completely possible and achievable—but
not in the Unicode Standard. It's out of scope for Unicode, it sounds
like. You've said you realize it won't happen in Unicode, but it still
can happen. Go forth and implement it, then: make your higher-level
protocol and show its usefulness and get the industry to use and honor
it because of how handy it is, and best of luck with that.
~mark
On 7/3/19 2:22 PM, Ken Whistler via Unicode wrote:
On 7/3/2019 10:47 AM, Sławomir Osipiuk via Unicode wrote:
Is my idea impossible, useless, or contradictory? Not at all.
What you are proposing is in the realm of higher-level protocols.
You could develop such a protocol, and then write processes that
honored it, or try to convince others to write processes to honor it.
You could use PUA characters, or non-characters, or existing control
codes -- the implications for use of any of those would be slightly
different, in practice, but in any case would be an HLP.
But your idea is not a feasible part of the Unicode Standard. There
are no "discardable" characters in Unicode -- *by definition*. The
discussion of "ignorable" characters in the standard is nuanced and
complicated, because there are some characters which are carefully
designed to be transparent to some, well-specified processes, but not
to others. But no characters in the standard are (or can be) ignorable
by *all* processes, nor can a "discardable" character ever be defined
as part of the standard.
The fact that there are a myriad of processes implemented (and
distributed who knows where) that do 7-bit ASCII (or 8-bit 8859-1)
conversion to/from UTF-16 by integral type conversion is a simple
existence proof that U+000F is never, ever, ever, ever going to be
defined to be "discardable" in the Unicode Standard.
--Ken
