The San Francisco Project

This is one of my favorite topics. The technology from the SF project has
come down through several generations since "Pink" was announced (I think
in '92). IBM and Apple were going to develop an object based operating
system for PCs to compete with DOS. Hewlett Packard even kicked in a
billion. 

But IBM and Apple figured out that it didn't matter how good an operating
system you made, there wasn't any way of getting it bundled with PCs. So,
the vision of Taligent changed to developing object frameworks for
application development. These frameworks were to be developed in C++ and
Smalltalk. The idea (for IBM anyway) being that such frameworks might
entice developers to create applications using new technologies that
access data and/or run on host based systems (normally thought to be Cobol
or RPG based). Work was going along well and VisualAge C++ and VisualAge
for Smalltalk were great development tools, when IBM hit the lotto.

Sun's intro of Java into the marketplace and the Internet acceptance of
the technology put a whole new spin on the Taligent frameworks. Now the
frameworks could be developed once and support all of IBM's platforms. 

The San Francisco Project is the development of those frameworks for Java. 

Now, what exactly is an object based application framework?

Well, as you know there are certain functions that all businesses need to
perform to survive. General Ledger, Payroll, Accounts Payable, Accounts
Receivable, etc. Let's take GL as an example.

This would mean that all object oriented business software must contain a
"dollar amount" object definition. An "account" definition, "credit",
"debit". A date, posted date, transaction data, due date, etc. Everyone
will need to post balanced transactions and run P and L statements. Read
records, process, print fields on screens, print fields on reports, format
dates, format amounts, etc. So there is a large amount of code that is
duplicated at each different site, or with each different software
package. Not all of it can be eliminated, but IBM has determined that they
might be able to eliminate 40% of an application development effort by
providing a uniform framework. 

A company like JD Edwards that has a lot of expertise in dealing with
certain market segments might want to generate an application that is
easier to maintain than procedural RPG has been. An application with a GUI
front end, network connections, and platform independance. But the severe
cost of redevelopment has kept them sitting on RPG code. Perhaps
eliminating 40% of the redevelopment effort is enticing enough to get them
to move. 

IBM's benefit from this is that these new applications can revitalize
their hardware. The software developer can re engineer their package to
appeal to new markets without abandoning their existing core customers.
Who could possibly lose from such an arrangement? ;-)

IBM has stated their current plan is to market their frameworks for a
percentage of sales. I suppose that means if you want to develop a public
domain package you can use them for free. 

I think there is more to consider and that might be discussed on this, but
I have got to get to bed!


Chris Rehm
Mr.AS400@ibm.net
You have to ask yourself, "How often can I afford to be unexpectedly out of 
business?" 
Get an AS/400.
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to "JAVA400-L@midrange.com".
| To unsubscribe from this list send email to MAJORDOMO@midrange.com
|    and specify 'unsubscribe JAVA400-L' in the body of your message.
| Questions should be directed to the list owner/operator: david@midrange.com
+---