dbwrapper/dbpgsql/dbpgsqlgather.. This is a database wrapper so we can
intergrate gnome-db into the system without problems.  This is handled by a
#define for database backend at compile time.  The dbpgsqlgather is specific
to whatever database backend your using and how to get that data from the
database.

dbobject is completely redone from scratch. A lot of functions were dumped
but still a lot of other functions need writing.  This depends on a lot of
the new functions. Big differences are it has a store of ids for fields its
read from and also the write cache is built into the object. (reverse of how
it operated before).  A lot of new structures here and a lot of stuff still
to be written.  file is a lot smaller now.

dbuniqueid, a new function. This is for id's. So no longer is everything
done on postgresql oid field, but handled by this file so it'll work with
other database backends.

db, I've restructued the API.  This wil be the main api file. new functions
like db_sql replace db_specialsql. db_getvalue replaces db_getfieldstr.  You
no longer need to create an object before you use it, and table name is
optional. however you need to specify tablename in order to write a new
object back to database. (you do need to know what table your saving to).
Also another thing is that db_getvalue get will tie in with a lot of other
functions directly like db_getdefaults etc.

dbbucracatt... (not spelt correclty) this is a global structure of every
database object created. Nice to know for memory links. But also to allow us
to know when object changes it effects another object so update its data

dbtoliet.  This takes over a lot of the dbobject functions.  This will do
the write backs and record where changes occur. though the changes are now
stored directly in a database object and not in a sperate write cache.  

dblog. This is a log of every object that changes. Useful for user interface
so it knows when to refresh widgets.

dboopath is renamed to dbpath

dbobjectvalid needs all its rules completely re-thought though cause of the
change in processes.

dbobjectbirth will mark how an object came into existance so it can be used
for checking for data effecting data.  

I'm hoping all this is going to be thread safe but i dont know till i try
cause tahts the next thing thats happening.  A number of the new files will
be running in completely different threads. Like writing to database will be
in different thread to reading, checking for log changes will be in another
thread as well..  hmm lots of threading..
