Tim

1) If I ruled to create my tables with *ALL* string fields with collate
"PTB_CI" (I understand it is Brazilian Portuguese) is there any problem with
performance ?

2) According with 1) the comparation with SQL does not need to use UPPER
neighter NOCASE. Am I right ?

3) The same things (as item 2) is applied to ORDER BY and SENSITIVE result
sets ?

Eduardo

Eduardo,

<< 1) If I ruled to create my tables with *ALL* string fields with collate
"PTB_CI" (I understand it is Brazilian Portuguese) is there any problem with
performance ? >>

No, there is really not much difference in EDB between the ANSI collation
and the language-specific collations, in terms of performance.

<< 2) According with 1) the comparation with SQL does not need to use UPPER
neighter NOCASE. Am I right ? >>

Correct.  Any indexes will inherit the CI flag by default, and any
comparisons will use the CI flag implicitly to do a case-insensitive
comparison.

<< 3) The same things (as item 2) is applied to ORDER BY and SENSITIVE
result sets ? >>

Yes.

--
Tim Young
Elevate Software
www.elevatesoft.com