Notes and Domino can be accurately described, in my opinion, as a "legacy" technology. I fully expect a similar pattern of responses from people whose jobs depend on Lotus Notes and Domino followed by it inevitably, eventually, becoming undeniable that I was right.

IT customers need to understand that what is more insidious than their own potential for denial are technology vendors whose products are obsolete. I personally believe that vendors' first priority is to obfuscate the single most important fact about their technology so that they can continue selling it as long as possible. Obviously, this is not in the interest of their customers, but, it is business. When we consider whether a technology is obsolete or not, we have to form our own opinions.

Let's look at the business drivers for SaaS. They include falling costs and increasing capacity as well as reliability of bandwidth, cheap compute capacity, rising data center operating costs (such as power and cooling costs), and the ever-growing burden of the supporting infrastructure required for messaging and collaboration (archiving, anti-spam, antivirus, regulatory compliance, legal discovery, and so on).

In my view, the underlying economic model upon which distributed client/server systems were based back in the 1980s and 1990s is dead.

TCO
The actual TCO of owning and operating an in-house messaging and collaboration system is not economically rational for most businesses today. Whatever top analysts say (and they disagree), Lotus Notes and Domino is a relatively expensive messaging and collaboration solution compared to the costs of ISP or ISP-like solutions. In my opinion, the cost simply cannot be justified in today's economic climate which continues to evolve in favor of SaaS.

There are many reasons why Lotus Notes and Domino is -- or once was -- an excellent choice (security for example) but the bottom line business issue is cost. It doesn't matter that Domino was the first application server in history because it's not a modern application server and doesn't scale comparably to products like IBM WebSphere.

Technology drivers
Now let's look at the technology drivers. Consider fast, reliable networks, multi-core processors with ever increasing compute density per chip; cheap, big storage and computer memories, and ever higher bandwidth. What makes sense today is not unlike what made sense yesterday: use the fewest number of servers for the greatest number of users optimizing for the lowest operating cost balancing line items like bandwidth and administration staff overhead (low TCO usually means high ROI).

In contrast, today's economics suggest that one server can and should service an entire enterprise with perhaps thousands of users around the world, or that one server can and should service many companies at the same time.

This is a new economic model that mandates ISP or ISP-like systems, not the distributed systems of the past. It means sharing the ever increasing gains in computing power and capacity across multiple companies so that costs go down.