Funny odbc + MS odbc.net behaviour.

Hi,

I have funny odbc (unicode) + MS odbc.net data provider behaviour, it's hard
to explain but I'll try.

Table structure which causes error:

CREATE TABLE test (
nimetus varchar(20) NOT NULL,
hind decimal(18,5) NOT NULL
);

Data in table

insert into test values('vvv vvvX',3); - X is char(382), Z with ROOF. Other
unicode chars provide same result. Eg. russian 'k'.

I have table with following row. I just make select * from test with
odbc.net provider and 'Input string isn't in a correct format' error is
arised.

Now funn begins:
If to change column 'hind' to decimal(18.3) - all works ok.

If to change row value from 'vvv vvvX',3 to vvvX,3 or vX,3 or v vvvX,3
... - all works ok.
(It's important that value is
char_char_cahr_space_char_char_char_UnicodeChar to get error, but not all
unicode chars won't fit)

Now I thought that's MS odbc.net provider bug, but to make same table with
same values to ms sql server using ms sql server odbc driver, all works ok.

pgsql 7.3.1 on windows + latest odbc driver

Have anybody some idea or suggestions ?

begin 666 bug2.sql
M[[N_#0H-"D-214%412!404),12!T97-T("@-"@EN:6UE='5S('9A<F-H87(H
M,C I($Y/5"!.54Q,+ T*"6AI;F0(at)9&5C:6UA;"@Q."PU*2!(dot)3U0(at)3E5,3 T*
M*3L-"@T*#0II;G-E<G0@:6YT;R!T97-T('9A;'5E<R(at)G=G9V('9V=L6^)RPS
$*3L-"@``
`
end