""Ann W. Harrison"" <aharrison@...> wrote in message
news:5.1.0.14.2.20020320121308.0244ea80@......

>
> The situation is worse than that. If you use a non-binary collating
> sequence - one that sorts upper and lower case in a rational fashion,
> for example, you increase the key representation. If you use a
> collating sequence that includes fancy things like diacritical marks,
> you increase the key representation again. So even if you use a
> one byte character set, you will be limited to 84 characters if you
> specify a collating sequence like FR-FR.
>
> And it's worse than that, still. If you have a compound key, each
> key is rounded up to five bytes and a tag bit is included every five
> bytes.

And the situation is better if you use a case-insensitive collation driver,
because you need less space in the index to pack each entry.

C.
--
Claudio Valderrama C. - http://www.cvalde.com - http://www.firebirdSql.org
Independent developer
Owner of the Interbase® WebRing