Re: RunSQLStmt in production

Although I once thought I'd like RUNSQLSTM to be able to run SELECT statements, I'm less inclined to that.

Because RUNSQLSTM is source-based, it is well-suited, as you say, to creating objects - just as DDS and RPGLE and other source is.

Now a naked SELECT? I find it to have been invaluable, especially when paired with output options like *PRINT and *OUTFILE. The several QMQRY-based EXCSQL/RUNSQL commands have always been able to run SELECTs. And the workarounds of CREATE TABLE AS SELECT... still don't allow for print output.

RJS has commands like this in several products - not a <verndor response>, just statement of fact. And they were able to run any command string directly entered, as well as use a source member.

Of course, all these QMQRY-based tools can do only one statement at a time. RUNSQLSTM can do multiple statements - again, making it excellent for building things.

Regards
Vern

On 3/29/2013 10:17 AM, James H. H. Lampert wrote:

On 3/29/13 5:31 AM, Robert Clay wrote:

The consensus seems to be that RunSQLStmt is primarily a
developer's/development tool and not necessarily designed for inclusion
in production processes. Agreed?

What could possibly have given you that idea?

Can you suggest a better way to automate the creation of SQL tables and
views, than to call RUNSQLSTM from a CL program?

Conversely, how would adding the ability to process a naked SELECT (as
opposed to one encapsulated in the creation of a VIEW) be useful in
RUNSQLSTM, given that we have STRSQL and various iNav functions (not to
mention freeware utilities like SQUIRREL) for that?

--
JHHL