RE: webfacing and competing products

["Nathan M. Andelin" <nandelin@xxxxxxxxxxxxxxxxxxx>]

> from: "Joe Pluta" <joepluta@xxxxxxxxxxxxxxxxx>
> I'll add that they're all a technological dead end, because they
> rely on the 5250 data stream.  That's what makes PSC/400 so
> much different and in my mind so superior.

I agree that PCS/400 is different.  Screen scrapers extract data from
pre-formatted streams before applying their own formatting, while PCS
essentially transforms record buffers BEFORE any other formatting has
occurred.

Perhaps Trevor's point was that a *DSPF still defines the formatting, and
that the user interface is still tied to the original *DSPF.  Actually,
Trevor seems to be struggling with understanding "thin" vs. "thick" client
architecture.  The assertion that Servlets are "thick" seems totally
off-base.

My feeling is that screen scraping may be strategic to the extend that 5250
is still strategic.  The biggest advantage of screen scraping is support for
IBM menus and commands, which are so tightly integrated into business
applications.

A large number of our software support calls were answered with instructions
about using WRKWTR, WRKOUTQ, WRKSPLF, WRKCFGSTS, WRKUSRPRF, etc. which our
applications heavily relied on.  Anyone deploying an alternative to
screen-scraping has those issues to deal with.

Yet, if the strategic goal is to move applications toward the
Model-View-Controller design pattern, while adding point-and-click
navigation to the user interface, then it seems to me that any tool
dependent on existing "display" files is probably a temporary solution.

Nathan M. Andelin
www.relational-data.com