On Tue, 2005-03-15 at 08:15, Helen Borrie wrote:
>
>
> At 07:39 AM 15/03/2005 -0600, you wrote:
>
> >I'll give it a go when I have a chance.
> >
> >The 25x byte count limit is a function of the engine, though. SQL, per
> >se, does not define a limit (although Firebird SQL may enforce one).
>
> Indeed it does. Ivan is right about the limit being 252, not 253. Ivan is
> the arch-guru on indexing, btw. :-)
>
>
> >VARCHAR(1) is 1 to 2 bytes. One byte for the length, and 1 byte for the
> >(optional) data.
>
> VARCHAR(1) stores a string of not more than one *character*. Varchars are
> stored with 2 extra bytes, in which a 16-bit integer stores the character
> length of the varchar. Therefore, a VARCHAR(1), stores zero or one
> characters, which might be one, two or three bytes, depending on charset,
> plus the two attribute bytes. It's the data bytes that are counted for
> index length, not the attribute bytes.
>
>
> >Try with CHAR(x) instead of VARCHAR(x).
>
> Don't bother. CHARs have the same limitation; and a CHAR(252) isn't a hot
> place to store a string of that size, unless you *really* have to store
> fixed-length strings of that magnitude.
>
> Anyway, the good news is that Firebird 2 lets you store ridiculously wide
> indexes - up to the page size, minus 4, minus (I suppose) a few other bytes
> that are earmarked for page pointers.
>
> ./heLen
>
>
>
>
>
> Yahoo! Groups Links
>
>
>
>
>
>
>