On Thu, Jul 28, 2016 at 2:33 PM, Jon Paris <jon.paris@xxxxxxxxxxxxxx> wrote:
[Aaron wrote:]
"My end goal is to make the experience rich/specific/restricted so the need
for a 5250 session would only be required if you are actually doing green
screen programming. “
I can think of many reasons for wanting 5250 that have nothing to do wth green screen programming - such as building and deploying an RPG program as a web service or stored procedure. Sure you could keep building tools ad nauseum to provide for this kind of stuff but haven’t you got better things to do?
I believe it boils down to needing **EASY** access to the native,
QSYS.LIB environment, because that's where a lot of important i stuff
lives. It's certainly where the stuff that is unique to the i lives.
(If you don't need that stuff, then quite honestly, you might as well
not bother with the i at all.)
Now, it just so happens that for most people, 5250 is by far the most
straightforward and reliable way to access the QSYS.LIB environment.
The Web-service-in-RPG example isn't much of a case for *5250*. It's a
case for QSYS.LIB. (Don't you basically bypass 5250 already if you
develop your RPG Web services and stored procedures using RDi?)
For me, even without touching RPG at all, I still would want easy
QSYS.LIB access to check on things like whether my Python program is
writing the correct data to a data area, or whether it successfully
submitted a native job. My favorite "hello world" test to confirm that
an outside program has successfully communicated with the native i
side is to send myself a message using DSPMSG. (And once this is
established, I can use it as a debugging tool when I try more
complicated things.)
I totally agree with the "but haven't you got better things to do?"
part. Strictly speaking, I don't need 5250 or green screen. But given
the alternatives available *right now*, it looks to me as though it
will be easier and quicker to figure out the configuration and
security issues surrounding bona fide 5250 access (to Aaron's
satisfaction for Litmis Spaces) than to build up 5250-bypassing,
QSYS.LIB-accessing infrastructure that provides adequate ease of use
(to my satisfaction for Litmis Spaces).
John Y.
midrange.com
