Subject | Re: [firebird-support] Remove duplicate of table without constraints |
---|---|
Author | Ann Harrison |
Post date | 2011-09-13T20:53:43Z |
On Tue, Sep 13, 2011 at 3:45 PM, Leonardo Carneiro
<chesterman86@...>wrote:
identified by column values. However, you can get the RDB$DB_KEY which is
sort of an id. However the RDB$DB_KEY is stable only for a single
transaction. Between transactions someone could delete one tuple with a
specific RDB$DB_KEY and another transaction could insert a new tuple that
could get the same RDB$DB_KEY value. Backing up and restoring a database
completely scrambles the RDB$DB_KEY values. Use a primary key.
Good luck,
Ann
[Non-text portions of this message have been removed]
<chesterman86@...>wrote:
>The id of a tuple in a table is not a relational concept. Tuples are
> How do i get the id of the tuple of a table? I mean the ID of the tuple in
> a
> table, not the PK of the tuple. Have i jumped any doc that describe this?
>
>
identified by column values. However, you can get the RDB$DB_KEY which is
sort of an id. However the RDB$DB_KEY is stable only for a single
transaction. Between transactions someone could delete one tuple with a
specific RDB$DB_KEY and another transaction could insert a new tuple that
could get the same RDB$DB_KEY value. Backing up and restoring a database
completely scrambles the RDB$DB_KEY values. Use a primary key.
Good luck,
Ann
[Non-text portions of this message have been removed]