In this 2006 article Nick Carr lays out the history of office productivity applications, arguing the Office 2.0 is really Office 3.0 - the generation where desktop productivity office suites mesh with the Web.

This article is linked to The Office question,
December 18, 2007

Thanks for the insightful commentary
Joe. I see things a bit differently. Maybe my tin foil hat is
wearing a bit tight these days, but i see MSOffice XML (MOOXML and
the MOOXML binary InfoSet) as a very important aspect of how
Microsoft integrates and leverages their desktop office monopoly
power into server side and device systems.

It is the combination of MOOXML and
.NET that creates the integration mesh between desktop, server
systems, and devices. Imagine every application or service
participating in either a loosely coupled or carefully crafted
information processing chain, being fluent in MOOXML, and able to
process internal data structures and processing instructions unique
to .NET.

Enterprise systems and services from
ORACLE, IBM and SAP will not have this same integration fluency.

The design of ISO MOOXML is such that
it would be impossible for <b>non Microsoft server and device
systems</b> to match the quality and depth of integration with
the 500 million desktops running MSOffice bound business processes.

Given that MOOXML will probably succeed
at getting ISO/IEC approval, removing the last "legal"
barrier for this MOOXML Stack, were looking at a massive migration of
MSOffice bound workgroup - workflow business processes to a new
lockin point; The Exchange/SharePoint Hub.

With the real estate industry, this
migration to to E/S hosted applications only took six months to
completely replace years of desktop productivity shrinkware
dominance. The leap in productivity was spectacular.

The downside of this migration is that
the real estate industry is now tied into Microsoft at the critically
important business process level. A binding that will perhaps last
through the next fifteen years. Try to take this away from the
average Realtor though, and they will bite your hand off.

The productivity payoff of this
migration is so dramatic that on one Friday afternoon, going into a
weekend of showing homes and submitting offers - the stuff that will
make or break every Realtor - Microsoft issued a summary end of life
for all Win2K desktops. And pulled it off without a hitch.

What MS did was to issue a security
patch for all Exchange systems. The patched systems then required IE
7.0. There is no IE 7.0 for Win2K desktops - notebooks. End of
Life. And not one complaint. The entire real estate industry simply
ran out to purchase new XP MSOffice 2003 systems able to run IE 7.0.
The lesson i took away from this is that the productive value of
these E/S Hub applications is such that writing a check to continue
business was seen as a requirement equivalent to having a license to
practice or a car to drive around in. It's a basic business expense.
They were happy to write these checks.

I believe that the MOOXML Stack is
designed to lock out competitors at a level they've never thought
possible. And i think it will work given that there is no such thing
as an alternative ODF Stack.

In 2004, washed away by the DOJ anti
trust settlement, Scott McNeely announced the terms of Sun's
surrender to Microsoft. He pointed out that Sun had find a way to
work with Microsoft because the Sun's customers were demanding better
integration of server side systems with the Microsoft desktop.

Sun sales people had tried to convince
server side customers that they could achieve the needed level of
desktop integration by replacing MSOffice with the inexpensive
StarOffice or the freely available OpenOffice. When that didn't
work, the sales staff stopped trying to bundle StarOffice with their
server side systems because it was killing sales.

Now Microsoft has their own server side
systems coming down the pike in droves. These systems all offer
superior integration with MSOffice productivity environments. They
also provide an easy to walk path towards the incredible productivity
jump a business process takes by moving from desktop bound workgroups
to an XML based integration Hub.

Server side vendors like IBM, Oracle,
Zimbra, Alfresco, RedHat - JBoss, SalesForce.com and Google are never
going to get that same kind of integration to the MSOffice bound
desktop.

When the server side guys competed
against each other, currying favor with Redmond was the secret to
marketplace advantage. Now that Microsoft has competitive offerings,
those favors might not be available. If the purchase decision is
going to be based on integration with the MSOffice bound desktop,
we're looking at a whole new monopoly realm of desktop to server to
device systems.

There's one last hurdle for Redmond.
Get MOOXML through ISO/IEC.

~ge~