Re: what was the single mgmt decision made in mid-1990s that slowed the evolution of ILE/RPG?

Steve Richter wrote:

On Thu, Feb 26, 2009 at 8:42 PM, David Gibbs <david@xxxxxxxxxxxx>
wrote:

(I know I'm going to regret this)

now you are channeling Jon Paris

I am mere padawan to Master Paris.

CL also is the baseline from which programs are called. Pgm A is
called, it returns an object, that object is then passed in a call to
Pgm B. So definitely, a control language has to be able to
instantiate and work with the data type of the variables of all the
programs it can call.

Although my experience is limited in this respect, but I have never seen a control language that could do anything like that.

A control language should setup the environment a program runs in and call the first level program ... at that point it's up to the application code.

Using the same arguments you've made, one could say that unix shell
script and windows batch files should be able to directly
instantiate and interact with Java objects (without needing to run
the java command).

exactly.

Ok, so ALL control languages are defective ... or, perhaps, your idea of what a control language does is wrong.

RPG cant use an SQL result set.

Sure it can ... that's what RPG SQL is all about.

nowhere to the degree it should be able to.

But you said that RPG _CAN'T_ use a result set. You didn't say it couldn't use a result set to a specific degree.

As I don't use SQL in RPG that often, I can't address specifics.

All of this is the responsibility of the run time framework,
which on the AS400 is/should be ILE.

Well, as Hans (welcome back, btw) has pointed out ... ILE isn't a
runtime environment. OS/400 is the runtime environment.

I dont follow why so many allow terms to bracket their thinking.

Bracketing thinking? Han's stated a fact. ILE was and is intended to do a specific thing. You are stating that it should do something totally different.

david