varChar field endian problem with Intel Macs

[jda]

Hi Ruslan,

I have users with a database made on a PPC Mac with an older version 
of Valentina (2.x). There is a varchar field, encoded as UTF-16.

When the db is opened on a PPC Mac all is well.

When it is opened on an Intel Mac, it seems OK, too. The data in the 
field are read correctly.

But when the data are edited and saved, when read back in they come 
out as garbage. You told me this was an encoding endian problem that 
was corrected in more recent versoins of Valentina, which is OK.

My question is, can I fix the problem programmatically when the db is 
opened for the first time with V3.x? That is, is there an SQL (or 
API) way to fix the field so that the user doesn't have what they 
think the db is corrupted?

Thanks,

Jon