It would be nice if Postgresql supported multi-octet raw data. Certainly a
lot of what you would do with it would be similar to bytea, but the basic
string functions would be overloaded so that the unit of work would be a
multi-octet word.
Multi-octet instances could be cast to bytea when one wanted to work with the
data bit-by-bit instead of word-by-word.
the declaration would be something like:
CREATE TABLE acme( ... ,multi_octet_foo MULTIOCTET (octets-per-word, max-words-per-column) ,...)
To reuse a declaration you could create a domain.
MULTIOCTET columns could be loaded using octal, decimal, or hexadecimal text
input.
The real advantage of a multi-octet type would be for power users. The code
would be reusable. It would simplify tasks like creating the often requested
native support for UTF-16 and the less frequently requested UTF-32.
