Re: SQL source

[Vernon Hamberg <vhamberg@xxxxxxxxxxxxxxx>]

Hi Joe

I don't see an inconsistency - a VIEW encapsulates a SELECT statement 
that is run each time the VIEW is used, an INDEX has no statement to 
encapsulate. Originally an INDEX was defined only by the key columns, 
returning all columns. Now it is possible to define the returned columns 
and put selectivity on it. I think, IIRC, that the key "column" can be 
an expression, but I've not checked on that.

So I see them as functionally different - a VIEW doesn't have an access 
path defined, exactly - again, IIRC. the access path is set up as the 
view is run. Optimization is handled as always, so the cache might have 
what is needed.

Cheers
Vern

On 8/12/2020 7:54 AM, Joe Pluta wrote:
> I'm not requesting any change, just noting something I found.  I know 
> there are other techniques, but I was just pointing out what seemed to 
> be an inconsistency.
>
> What is consistent is that the SQL file type is always set for 
> SQL-generated logicals, so that can then point me to use one of the 
> other techniques that you and Birgitta referred to.
>
>
> On 8/11/2020 2:02 PM, Rob Berendt wrote:
> > Think "outside of 5250".
> > Instead of trying to look at DSPFD you can retrieve that information 
> > from SYSINDEXES.  Or you can look at it with ACS in the Database 
> > option.  See also GENERATE_SQL at
> > https://www.ibm.com/support/knowledgecenter/ssw_ibm_i_74/rzajq/rzajqprocgensql.htm 
> >
> >
> > But if you want to submit a RFE to get that added, feel free.
> >
> > Rob Berendt