FreeTDS Development Group

[Frediano Ziglio <freddyz77 AT tin.it>]

> 
> 2. There is a uniqueidentifier patch ready for DBD::Sybase, but until then
> it won't handle that datatype.  Need UG update on that.
> 

I fixed ct_fetch. It return CS_ROW_FAIL if a conversion error is
detected and set correctly indicators...

> 3. tds_convert() shouldn't issue an error for anticipated zero-length
> conversions.
> 

dbconvert and cs_convert already check if srclen == 0 and do the right
things... I think odbc and tsql should be changed to this behaviour
too... But not so sure... What's your opinion ?

> 5. Bad uniqueidentifier handling in sqsh is probably due to sqsh not
> understanding the datatype.  
> 

I think so, tsql do not have problems...

> 6. Eric experienced problems with DBD::ODBC.  I consider that acceptable for
> 0.60.  Nothing personal, 
> Eric!  :)  I don't want to hold up the release for a problem that might take
> a while to isolate.  If OTOH it is isolated and can be fixed quickly, then
> of course we should include the fix.  
> 

I think odbc is a bit broken... As Bill test show explicitly...

> Brian alleged "spurious conversion errors".  Anything specific/confirmed?
> 
> No new bugs.  Is good.  Expect -rc1 on Thursday or Friday if all stays quiet
> on the western front. 
> 

It seem only a NULL problem...

freddy77