Indexing and collation question

[jda]

Hi Ruslan,

My db and its indexes are now collated as 
kSecondary (searches are diacritic sensitive).

Some of my users are asking for 
diacritic-insensitive searches. Since I want 
indexed searches to be affected only, I thought 
that I'd add this to the main index (allFields).

allFields.CollationAttribute(EVColAttribute.kStrength) 
= EVColAttributeValue.kPrimary

This seems to work just fine and now searches for 
u also find ü (umlaut over the u).

I have to questions:

1. Is this safe? Am I missing something obvious that might bite me later?

2. Since I'm using the classes way, I'm assuming 
that this can't be changed once a db is created.

Thanks,

Jon