Re: SQL question on join and limit rows retrieved

[CRPence <CRPbottle@xxxxxxxxx>]

Porterfield, Sean wrote:
> I have an RPGLE program that works and lists header/detail (one
> to many) for selection.  This means a standard SETLL/READ or
> READE on the header and CHAIN to detail once to get the first
> record.  The remaining detail records are only retrieved when one
> entry is selected.  (The first detail record is intended as an
> overview / brief glimpse.)
>
> I want to duplicate this with SQL to allow for easy enhancements
> and web enablement.
>
> <<SNIP>>

  Making the described happen in SQL is effectively a /trick/ [for 
lack of full details of the detail table; seems a reply already 
suggests there is no /line number/ for example] which will most 
likely produce _different_ [not necessarily undesirable] results. 
That, and probably with a higher cost in all of CPU, memory, and 
temporary storage.

  Given an already functional program, apparently already efficient 
with keyed selection, how is changing to SQL expected to effect both 
easier enhancements and web enablement?  Is the intent to perform 
the query from the web application rather than generating a result 
set by calling a stored procedure?

Regards, Chuck