Re: Oracle reports: linked queries

[Quoted] By using independent queries and joining them you gain the ability to have each group/query behave differently on the layout. Such as.... if you want the first group to repeat 'down' and the next group to repeat 'across and then down.'

Did you include your 'order by' in you query? Remember that your break groups effect your sort order also. Also, having independent queries can resolve the need to have outer joins on table that are 'outer joined' (which [Quoted] causes an error) in the one query. Each query group will have its own order. If an outer query needs to order according to columns in the first query you should include a reference to that table in the second query even though you may not redisplay those columns on your layout.

When I created a transcript for a university I used 43 queries that reference 90 tables or view and the whole thing fit on one page and was quite amazing to have each area such as 'Coursework Completed', 'Coursework in Progress', 'Attributes for Coursework Completed', 'Computer Competency Requirements Status' ,'Langueage Requirement Status', etc, each have their own behavior on the layout. It ran very fast, also. rq -- to resond by e-mail remove nospam from e-mail address.

Scot Balfour wrote:

> My team is trying to understand the differences between writing a single
> query that joins several tables versus and writing one query per table
> and linking them together.
>
> It seems like when you link the independent queries: A and B, you can't
> select the order of the fields that appear in the default layout. We
> wanted a report that included fields: A1, A2, B3, A3, B4 but the default
> layout always took all of A's fields first, then all of Bs'. We could
> omit fields, but not resequence.
>
> We concluded that if you needed to include fields from multiple tables
> and you wanted to order the default layout by alternating between fields
> from table A with table B, you needed to write a single query with a
> join, instead of linking independent queries.
>
> Can anyone validate this conclusion or better explain their
> success/rationale behind linked queries? Thanks.
Received on Wed Dec 09 1998 - 11:26:21 CET