Semper fidelis
Tim Bray: In the past few days I’ve been watching two debates on the subject of Unicode
This lead me to a morning full of exploration, where I learned quite a few thing about m17n, a term I had not come across before.
I was
aware that
the latin a
(a
) and cyrillic
а characters
(а
) are different characters, but display using the
same glyph.
I was
aware that
the latin
small letter e with acute could be encoded either as é (é
) or
é
(é
)
and that both encodings are intended to be canonically equivalent.
What was news to me is that characters like Unicode Han
Character ‘enter, come in(to), join’
(入
) have common semantics across various so-called CJK languages but displays
as 入 (zh-tw
) or as 入
(ko
) depending on the value of
lang
and/or xml:lang
, and that this
controversy
is one of several factors which are shaping Ruby’s
Unicode roadmap.
This apparently is very troubling to Tim who (like most of us) have
neither the inclination, time, or means to fork either the IETF or Ruby, which
leads me back to
Mark Pilgrim’s quest for fidelity that lead to the statement that
Openness is not a
cargo
cult. Some get it, some don’t. Apple doesn’t.
This sentiment seems to me to be the underlying foundation for Mark’s
beautiful theory that is increasingly at odds with uncomfortable facts. Ones like
Jacques’ observations on how to obtain the best fidelity for MathML with various products.
A theory that I’m more comfortable with was expressed over half a decade ago in Mark Pilgrim’s Misspent youth. Generalizing this, it seems to me that:
- having multiple independent and demonstrably interoperable implementations of a given standard is the most important thing
- having at least one of them be open source is arguably the second
- actually using an open (or Free, if you prefer) is a distant third
Unfortunately, this theory has to contend with its own set of ugly facts, namely Perl, PHP, Python, and Ruby.
Bah