RE: "Giga Character Set": Anything besides noise

From: Kenneth Whistler (
Date: Wed Oct 11 2000 - 18:08:28 EDT

I'd have to agree with Carl's assessment.

Digging a little further, I note that:

Coventive Technologies is a Taiwan-based Linux company founded in
September, 1998, according to their web site (which doesn't scroll
pages correctly, either with IE or NN, by the way). Hmm. Isn't Taipei
the home of the "better" Chinese character encoding-of-the-month

The announcement on, copied from the white paper
posted on is so riddled with errors and
misconceptions that it is hard to figure out exactly what they are trying
to say, except that GCS is better than Unicode, so you should buy their
Linux systems. Clearly, however, they are not promoting a *standard*, but
rather a proprietary (and patented--they say) internal processing system that
is supposedly more efficient and faster than Unicode -- though their claims
appear to be nothing more than market hype. ("For example, GCS handles
Korean font files 1500 times faster than Unicode..." Huh? I guess they
measured the file I/O in one case and not in the other.) But since they
are not promoting a standard for characters, that can only mean that
they are apparently depending on other people's standards for defining
the characters. And that means that their "multilingual" solution is
just code pages hiding behind a "mathematical encryption algorithm".

A quick search of U.S. patents turns up nothing for "Giga Character Set"
or "Coventive", so I cannot even evaluate the technical claim of the
patent -- if one does, in fact exist, as opposed to just a patent filing,
and that possibly (probably?) not even in the U.S.

John and Rick have already commented on the "sheesh!" factors about
many of the claims in the white paper.

I'll be sure to stay out of the way of the stampede of eager customers
chomping at the bit to buy into a proprietary "display code" approach
with completely undefined interoperability with anybody else's data
or for that matter with the Internet itself, just because these
guys assure it that it runs real fast and that unlike Unicode, it
"preserve[s] the characteristics of language."


> David,
> I have not but it seems like they are hawking it because they own the
> patents on it and want to make money. But with "A different algorithm is
> developed for each language" reminding me of the horror of codepages and
> "GCS 'calculates' or derives the correct character and font" brings to mind
> a glyph oriented system, it seems that GCS will be another DOA idea.
> Carl

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:14 EDT