Some more info on this,

If I use IBConsole to look at the data it works OK, and I can see the data
in the table until I scroll down to one of the new entries. Then it beeps
and a dialog pops up that says "Invalid Data Conversion" and the table goes
blank.

JayBird is perhaps storing dialect 3 timestamps in a dialect 1 database? I
don't know enough about that mechanism to be able to tell.

Oh well, I needed to convert those DB's to dialect 3 anyway!

Rick

----- Original Message -----

> I was completely wrong about this, Blas has simply moved the code
> elsewhere. I don't know what the problem is.
>
> Sorry for the false alarm
>
> david jencks
>
> On 2002.10.31 12:33:13 -0500 David Jencks wrote:
> > Hi,
> > this is more or less my fault. For a couple of days I had checked in by
> > mistake some experimental code to try to identify the source of some
date
> > problems (the encode/decode date routines are slightly broken). Blas
> > seems to have worked from this version with his performance
improvements.
> >
> > I should be able to fix this shortly.
> >
> > thanks
> > david jencks
> >
> > On 2002.10.31 11:31:57 -0500 Rick Fincher wrote:
> > > Hi All,
> > >
> > > I just updated an older program that uses a dialect 1 database with
> > > yesterday's cvs code.
> > >
> > > The date behavior was strange. The program didn't recognize any of
the
> > > dates inserted as timestamps into the database by the old program,
but
> > > worked OK with new insertions.
> > >
> > > When I looked at the database with FBWorkbench (which recognizes it as
> > a
> > > dialect 1 database) all the old dates appeared normal but the new
> > > insertions
> > > had dates in 1859, even though the Java program with the new JayBird
> > was
> > > interpreting the dates correctly.
> > >
> > > The older JayBird beta driver worked fine with the old dates.
> > >
> > > Any ideas?
> > >
> > > Rick
> > >