🚀 go-pugleaf

Markku Sakkinen <sakkinen@jyu.fi> writes:
|
|   I don't think that the stupid idiosyncrasies of one particular
|   programming language should be taken into account when a long-time
|   universal character code is defined.

There are such things as "installed base" to consider when doing things.
Although I share your emotional reaction, I also know that standards
that change too much of the way things work will never fly.  I'd like to
see ISO 10646 fly.

I don't know whether this is relevant, but ISO 10646 suffers from "byte
order" problems when it's a word or half-word (word being 32 bits) code,
whereas well-known multiple-octet coded character sets such as JIS X0208
define a byte order.  ISO 10646 has a "signature" (FEFF) and an unused
(or prohibited, actually) value which corresponds to the byte-swapped
signature (FFFE), and so allow brain-damaged hardware to look like real
hardware.  (Talk about idiosyncracies!)

Aside from the hardware problem which makes it impossible to view raw
ISO 10646 as a word or half-word code to portable C programs, the zero
terminator could be defined to be a word or half-word, and everybody
would be happy.

Best regards,
</Erik>
--
Erik Naggum             |  ISO  8879 SGML     |      +47 295 0313
                        |  ISO 10744 HyTime   |
<erik@naggum.no>        |  ISO 10646 UCS      |      Memento, terrigena.
<enag@ifi.uio.no>       |  ISO  9899 C        |      Memento, vita brevis.