Re: Re: Re: Modify the IE Browser

> Let's keep in mind that robust Web applications are
>stateful too.  The difference generally is in how state is retained.

Yes!  Much better than I said.

> My point is that both types of interfaces serve a useful purpose.
> The 5250 paradigm still makes sense.

Yes it does, and it's an interesting time trying to marry the RPG paradigm
to HTML.  You have said it better than I could.

As for Joe's comments regarding keeping the logic that's in place, I can't
agree more.  Yes, it sure seems like a temptation to completely re-write and
start over (now that I'm so much smarter than I was - HA!)  But look at
Netscape: whilst they re-wrote their browser to fit in the constrained
memory model, Microsoft blew them apart in the marketplace.  Very few ISVs
are willing to go down the same road, which is why 'refactoring' has such
appeal.  I can also speak from the home-grown software side of the house,
having done that for 17 years.  My employers were completely unwilling to
spend an hour to re-write something that was working well enough.  If I had
to go into the code for other reasons, it was OK to re-factor, but extensive
re-work was frowned upon.  The last major code construction/development
effort was back in the days when we moved from card/batch processing to
workstation/transaction at a time processing.  And even then, many of the
'individual transactions' simply fed our old batch files.  It took years of
incremental improvements to become completely transaction oriented.

The point of this monotonous note is to try to explain what companies with
small budgets and small staffs go through when faced with a major technology
change.  I hope I was able to convey some of that...

  --buck