Nicholas Bodley scripsit:

> Caveat: I have only scanned the document, but what I did see
> looks very sensible.

I have read it, and it is riddled with elementary errors and
misunderstandings about Unicode. Furthermore, it has many of the same
deficiencies it accuses Unicode of having, but in spades; for example,
each of the hundreds of languages that use the Latin script gets its
very own encoding, so A-Z and a-z are encoded hundreds of times. It has
nothing to say about bidi, multilingual sorting, or many other issues.

> Nevertheless, despite all the excellent, careful, thoughtful,
> heartfelt and informed work done to create Unicode, it does have some
> weaknesses.

Quite so. Unicode, like the scripts it encodes, is the product of a
historical evolution, and cannot be expected to be wart-free.

> In particular, some versions of a browser render Vietnamese text
> with the uniquely-VN letters displaced in both x and y, bold, and
> maybe a different point/pixel size. The rendering looks like something
> the cat dragged in, with hints of a ransom note. Switching fonts and
> weight/slant within a word just should not happen, unless the author
> explicitly intends it to happen.

This is not a deficiency in Unicode, but the browser's last-ditch attempt
to render characters not in its principal font using any available font.
If you install a properly tuned Vietnamese font, the text will look
correct. This is a question quite separate from that of character
encoding.

> I still love Unicode, but I also intend to study this document with
> considerable care.

IMNSHO it is a waste of time. Also note that it is dated 1997, and meanwhile
Unicode has taken over the computing universe.

--
John Cowan <jcowan@...>
http://www.ccil.org/~cowan http://www.reutershealth.com
Charles li reis, nostre emperesdre magnes,
Set anz totz pleinz ad ested in Espagnes.