Ok, I now understand more about how Postgres handles these older dates but
we're still seeing a problem that I'm not sure how to handle. Simply put, we
write a TIMESTAMP via JDBC and then read it back. What we write and what we
read are different. The only way I can think of fixing the problem is having
a check, in our code, to see if the dates are earlier than Postgres' magic
cutoff day, and, if so, do timezone fixes. This is bad. Is there a better
way. Is there, perhaps, a bug in the JDBC driver?
Thanks,
--Rainer