Re: some ct_options / ct_diag questions

[email protected] wrote:
> is the CT-library part of freetds supposed to exhibit 
> exactly the same behaviour as the standard Sybase CT-lib? 

Yes.  ct-lib's error reporting is not as good as that of the other
libraries, though, owing to previously described lack of developer
experience.  ;-) I'm not surprised the error handler isn't invoked when it
should be sometimes.  

> Sybase CT-lib reports a row count of 0 
> and freetds CT-lib a row count of -1 (CS_NO_COUNT).
> Before I dive into the freetds ctlib and/or tds sources, do you have a 
> quick thought on why this happens?

The TDS prototocol returns a rowcount and a bit saying whether or not the
rowcount is valid.  Either ct-lib is ignoring the rowcount, or it's
reporting that the invalid bit is set.  Perhaps Sybase's implementation
relies on knowing  (based on the server message) that no rows were sent. 
Or maybe it just processes the datastream correctly.  

Something in that vicinity. 

HTH.  

--jkl