FreeTDS Development Group

[Casey Allen Shobe <cshobe AT osss.net>]

James K. Lowden wrote:
> Could you send a log file of a test case, please?

Not easily at this point, sorry.  But a bit of guesswork fixed the 
problem, so do you still need it?

> One working and one rejected insertion would be good.

They both looked identical - I saw the error about invalid cursor state 
regardless and the application acted like it was inserting with no 
errors, I simply noticed that the destination table didn't have any data 
in it.  Turns out I was inserting a timestamp(6) without time zone value 
form postgresql into a datetime field in SQL Server, which can only 
handle a timestamp(0) without time zone.  A cast in postgres corrected 
this.  Usually SQL server will give me some error about trying to insert 
the wrong datatype and my script will die at that point.

So I guess it's a different bug entirely, but here's an example of what 
I was trying to insert into a datetime column: '2004-03-05 10:34:03.638996'

- Casey