> Hi Jan,
>
> We had something similiar to this in the past. The entire text was broken
> down into lines (80 characters each) and stored as individual rows in a
> table. The problem with this approach was that you can't search for phrases
> that span rows, so I'd like to be able to store the entire text inside 1 row
> and be able to search it. Oracle has a LONG datatype which provides this kind
> of funcionality. I was hoping that PostgreSQL had something similiar. We are
> rewriting portions of our legacy system using Java and trying to provide as
> much portability to RDBMSs as possible. Although our main production machine
> is Oracle, we'd like to make this run under PostgreSQL as well. I guess our
> only option is to go back to breaking things up
> into lines again (atleast for the PostgreSQL implementation) and providing
> special code that works for this database. I just hate cluttering code with
> special conditions (i.e., if
> System.getProperty("DATABASE").equals("postgreSQL") then
> do the following).
> >
> > After 7.0 is out, I'll continue the TOAST project. This will
> > break the size limits for variable size attributes and make
> > them virtually unlimited (new limits will be based on
> > available per process memory, so U can create more swap to
> > increase it). Stay tuned!
If you don't mind to have a somewhat crippled DB schema
during one release (7.0 until TOAST), let's see how we can
simulate LONG columns using a set of rules and triggers. Wait
some hours and I'll tell ya how it could work - if what I
have in mind works at all :-).
Jan
--
#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me. #
#========================================= [EMAIL PROTECTED] (Jan Wieck) #
