FreeTDS Development Group

[Stefan Bodewig <bodewig AT bost.de>]

Hi Stefanos,

>>>>> "SK" == Stefanos Karasavvidis <stefos AT multimedia-sa.gr> writes:

 SK> The hole issue about the use of the unicode character types is to
 SK> work with unicode text.

I think Brian was referring to the TDS protocol not the C library or
the JDBC driver when he wrote:

 SK> Brian Bruns wrote:

 >>   nvarchar are automatically converted to varchar and truncated
 >>   (if necessary) when SQL Server is talking TDS 4.2

In fact - just experimenting a little - SQL Server 7 returns VARCHAR
columns (at least this is what it tells freetds_jdbc) when you request 
NVARCHAR values.

 SK> The freeTDS driver does correctly return the Strings even if
 SK> characters from different code pages are included.

But this is not really because you are using nchar columns I
think. Did you try your combination of characters with an ordinary
varchar column?

Then maybe you could try to use text instead of ntext - 
CONVERT(text, column). This really should not be the final solution
but might be the best you can get at the moment.

 SK> My problem is that I have to use these kind of fileds, so I
 SK> would appreciate if someone could tell me if there is some
 SK> schedule to support this.

As Brian said. You will need a TDS 7.0 capable JDBC driver which
doesn't exist yet. Don't know about a schedule, sorry.

Stefan