RE: 5250-to-HTML servlet?

> From: Price, Chris
>
> Joe,
>
> I'm guessing here, but I would have thought that since IBM have enabled
> Webfacing to run without the interactive tax on the new hardware, that the
> suppliers of similar products (such as Seagull, Jacada, and even Pluta
> Brothers) would (should?) be hammering on IBM's door very hard, to ensure
> that their products can take advantage of the same way of running (however
> it's technically achieved).

Well, Pluta Brothers (and others) have been asking for a workstation exit
point for some time now, but never really expected to get one.  It's a
revenue issue with IBM.  However, even if they do give us such a thing, it
will simply be yet one more thing that will help the intrusive products but
won't help the screen scraper folks, because they're too far down the
stream.  The screen scrapers all rely on the already formatted 5250 buffer,
whereas the API IBM is using must be at a lower level (basically
intercepting each I/O call).

For example, the exit point would be called on a WRITE to the subfile
control record that initializes the subfile, then on every WRITE to the
subfile, and then again by the EXFMT to the subfile control record - this is
exactly how PSC/400 works.  In the screen scraper paradigm, they never even
SEE all those writes; they simple see the first page of the display.

No, unless the scrapers want to fundamentally rewrite their products,
they're basically now all competing for that percentage of the audience that
can afford interactive.  I wouldn't want to be a screen scraper company this
year.


> I also think it would be a very bad move by IBM if they denied this to the
> ISV's. After all, their whole strategy is based on migrating 5250
> applications to the web, by whatever means necessary. And it's not as if
> they would lose sales revenue - they give webfacing away for free anyway.

Who knows what IBM thinks?  This is the same company that killed WAS35SE,
then brought you Tomcat, then killed Tomcat and brought you WAS5Express.
This is the same company that told you RPG programmers would be flipping
burgers, then added a flurry of wonderfully compatible improvements to the
language, then said the MOVE opcode is a bad thing.  IBM has not shown great
consideration of ISVs over the years.

Joe