Re: ODBC (was RE: Green screen - it's time is over )

----- Original Message -----
From: "Joe Pluta" <joepluta@PlutaBrothers.com>
To: <midrange-l@midrange.com>
Sent: Saturday, November 17, 2001 3:36 PM
Subject: RE: ODBC (was RE: Green screen - it's time is over )


> I decided, for the fnu of it, to take a look at what ECS does,
and then
> re-read your post to see if it made any more sense in
perspective.  In some
> ways, it does, and some ways it doesn't.  ECS stores and
retrieves
> documents.  It's a document retrieval system.  No business logic
whatsoever,
> although the documents stored are created by business logic.
This
> interesting dichotomy may be what leads you to some of your more
interesting
> statements.

Sure, we wrote the first client/server based report archiving
system for the AS/400, with a PC as a server and the AS/400 green
screen as the client. We did it over twinax, no PC Support or
Client Access, no ODBC. That was in 1990-1991 as I remember, We
used the AS/400 as the transport layer. It was fun. It was
actually a rework of a 1989 implementation I did on NCR
interactive systems. It performed better on the AS/400.

When TCP/IP became valaible on the AS/400, we rewrote our product
to support green screens via Telnet, as well as a Windows Client
and a web client. We also wrote a custom client for the AS/400
green screen to give better functionality than the AS/400 Telnet
client.

We also code our own database access for prospecting and customer
relations as a web server application. Originally it was done as
Cold Fusion, now it is compiled VB running under the ASP
framework. We've run our business on a web client for five or six
years. Our programmers figured out how to run VB in the
interactive environment while creating ActiveX DLLS to run with
IIS. Fast development, very fast execution.

I also have another software companuy that has an elearning LMS,
also web based.

I got my start as a software house writing complete application
systems - job cost, payroll, gl, ap, ar in COBOL. It took me the
better part of a year and a half. Then I wrote a source code
editor, then a macro precompiler. Then I wrote an English-language
query system (this was before SQL was available on minicomputers
in COBOL - the only language available.)

When PCs came in I wrote some terminal emulators in QBASIC plus
some assembler for screen handling and communications. I wrote an
NCR, TI, and what would now be called ANSI-standard emulators.

I wrote the first version of the archive application, in QB and
then VB. I also wrote the key access system, because PC databases
were too small at the time (and are still too slow, and leave the
key nodes half full.)

I also wrote the paged compression scheme that lets you randomly
position to the virtual location in the files and decompress 32K
pages - we needed compression in the early days when hard drives
were 40 MB and opticals were 200 MB. Up to 6X compression and
lightning fast in random access. A simple but rather elegant
solution.

I also wrote Picksee, a CGI-based (it was years ago) database
query system for the web.

We wrote our own programmable LPD and FTP to talk to the AS/400,
so we can program extra functions into them. We write to the
socket level. We also wrote a Carnivore-type  system that we were
about to release when the FBI version generated so much negative
publicity. Oh Well.

I wrote a distributed telemarketing and customer service system
that uses voice modems to dial and record outgoing calls and
upload them over the web to attach to the prospecting database, so
the sales rep or manager can review entire calls while the caller
works from home over the web. You can pay the agents as
independent contractors and pay them by time on the phone rather
than elapsed time.

Every one of our public applications has to stand in the
marketplace against some very eager and rough competition. That
adds an interesting dimension to the design and programming cycle.

As I said before I don't disagree with your notions entirely. I
think you can find some better arguments for selling your ideas to
people. We wrote our elearning system from scratch in the last two
years, and there is no logic in the client - it is just
presentation.