Re: IBM i "Open I/O" Architecture

Jon wrote:

What many people miss is that if the vast majority of RPGers
were going to write and use subprocedures for everything outside
of basic RPG I/O they would have done it by now. Lets face it,
subprocedures have been around for 10+ years and are still not
extensively used in more than perhaps 30% of shops. The reasons
why are many and varied but certainly include the fact that a
large number of programmers have continued in the belief that if
the web etc really mattered to them then IBM would have given them
op-codes. Open I/O is the closest to that option that we are going
to see - if it can't move those folks then there really is no hope.
But even for those of us who do use subprocedures, Open I/O still
offers some very interesting possibilities.

Jon: You've accused me of negativity in the recent past. But reading this
doesn't help make me feel more optimistic about RPG's future.

I'm sure you can't comment in any great detail since I'm sure you're under
NDA. But the rest of us are left to speculate on what "Open I/O" means for
RPG. The suggestion that this is all within RPG dampens my own over-active
imagination. This could have real potential with the right support from
the O/S, in my opinion.

The way I see it, if people expect this to be some sort of "magic bullet",
I suspect they will be disappointed. Here's one scenario I envision for
web apps: I see Open I/O as a means of abstracting the presentation layer
from the business logic. Rather than outputting raw HTML, you pass buffers
of data via an I/O-like interface to the program that formats the data.
This doesn't really make things any easier. It's just a different way to
structure the application.

My thoughts are still wavering back and forth on this, but I can't help
feeling this "Open I/O" is a bad direction for RPG. Look, for decades,
we've been telling people to restructure their apps to separate I/O from
business logic to prepare for new forms of UI. (You know, the old MVC
model.) This kind of thing may well encourage the embedding of user
interface logic with business logic.

Will this have any impact on the ~70% of shops that haven't even adopted
procedures? The current hype may get some of them to learn a bit more,
which would be good. Will it be enough to encourage them to embrace more
current RPG programming techniques? I suspect most of these programmers
will be doing RPG III programming until the end of their careers.

Cheers! Hans