Could someone clarify what is apropriate for these situations for me
Firstly I'm using 7.0.2 on Linux and FreeBSD and am in the development
cycle for a system.
My middle tier programmer asks me from time to time to add another
attribute to views that I have made, and everytime I do, I feel like
I'm tangling with disaster....
I have nested views and views which call on plpgsql functions,
if I happen to make a modification (using pgaccess) to a view which
has another view which uses it, it corrupts things and I have to drop
all of the views which make reference to the modified one and recreate
them... OR I can use pg_dump to allow me to make the modifications to
a dumpfile, drop the db and recreate using the modified dump.
Anyways, I am understanding the behaviour of this after fighting with
it for a while, but I'm wondering why it is so hard to do what I'm
doing... Are nested views a big no-no??? same deal with plpgsql
functions... if a view uses a plpgsql function and you make a change
to the function, the view won't work again until you drop it and
I've figured out that it has to do with creation/saving order for
visibility... in the lookup cache...
What SHOULD i be doing? Will 7.0.3 help me with this??? 7.1?
pgsql-novice by date
|Next:||From: Arcady Genkin||Date: 2001-03-10 11:44:37|
|Subject: Not incrementing the 'serial' counter on failed inserts.|
|Previous:||From: Charles Welch||Date: 2001-03-09 19:17:46|
|Subject: Number of Records in a Table|