This blog will be used to communicate important milestones and events affecting the software industry and this company.
Our Objective
There are lots of articles trying to promote cloud computing. Most often the term cloud computing is confused with “Software as a Service “ (SaaS) or “Infrastructure as a Service” (IaaS) which some companies are trying to sell as cloud computing.
Our cloud computing model is based on the concept of a geographical (or local if needed) computer cluster concept and will reside under full control of the customer’s data center personnel. Its cost will be very competitive with the SaaS and IaaS models. It will provide much higher reliability and security than SaaS or IaaS models.
How is this possible?
When we look at computer technology, it resembles a giant puzzle. Each piece of this puzzle is often developed independently of other pieces. Very few people really grasp the large picture and most of us work confined to a narrow field.
This is the reason why most improvements that we witness are inadequate. We can create better computer languages, better operating systems and better databases, but when it comes to creating a great product combining all of them, we have many problems.
Software Abandonment
Having the perspective of almost thirty years of commercial IT industry experience, I have been observing for a long time the interesting phenomenon of organizations underusing or walking away from expensive commercial software.
This interesting phenomenon started in the early nineties with the organizations abandoning very expensive COTS software. Even when software was not outright discarded, it was underused. This phenomenon continues and is becoming more expensive as the following article illustrates:
http://www.zdnet.com/blog/projectfai...e_skin;content
What is interesting about the problem described in that article is that the sum of money wasted was large enough so it could not be covered up.
The value of software being abandoned ranges from a few thousand to tens of millions of dollars. Most of it is small enough to escape headlines, yet it seriously affects the bottom line of the organizations experiencing it.
To use an interesting and well known imagery, this is the case where COTS Emperor has no clothes. And, as in that story, the fear of punishment or of looking incompetent provides the incentive for hiding and perpetuating this problem.
In most cases, articles and consultants point to the "usual suspects" as being the cause of this or that fiasco. The lack of communication, bad management and lack of attention to details is on the top of that list.
The real root causes are more complex and disturbing than these "usual suspects" and are not quoted in articles because they would question the integrity, quality and lack of innovation within the IT industry.
I see three culprits responsible for this phenomenon:
- (Obsolete) commercial software architecture and design
- Lack of expertise within end user IT organizations
- Lack of expertise within consulting organizations integrating these products for customers
1. Software architecture and design.
It is based on assumptions from the seventies and eighties. In that period there were few end-user IS organizations and most of them were technologically competent.
The junior personnel were mentored by senior personnel and this insured the propagation of technical expertise. These mentors were promoted to the management level and provided necessary technical leadership to their organizations. This model broke in the nineties. Penetration of computers into all aspects of corporate and private life created a demand for IS personnel that could not be met.
The management of IS organizations became non-technical, and the mentoring system broke down. Yet the software is still targeted at the same type of personnel as in earlier decades.
From my own experience, I see that the quality of what passes for commercial software was never as low as it is now. The software being sold as a solution is often just disguised middleware requiring lots of work to make it usable. The integration work needed to make such software usable often escalates to a full fledged development effort. This requires a tremendous amount of resources, and this brings us to the second point:
2. Lack of expertise within end-user IS organizations
Very few organizations have the expertise to integrate existing COTS software. Some of them struggle with relatively simple software (both COTS and open source).
The expectation for end-user organizations to maintain the expertise necessary to compensate for the shortcomings of the IT industry is unrealistic. Their focus should be on their own businesses, instead of being the permanent proving ground and source of financing for half-baked ideas and software thrown at them.
This situation creates interesting Rube Goldberg solutions defying common sense and brings us to the third point:
3. Lack of expertise within consulting organizations
There are many articles about the decline of technical competence in the end-user organizations. The remedy for this, from the end-user point of view, is to supplement their own expertise by hiring contractors and consultants.
The problem with this approach is that, from a small number of consulting and contracting organizations and individuals in the past, their numbers have expanded to hundreds of thousands, all competing in the same marketplace, offering the same type of services, methodology, and level of expertise.
The same personnel circulate among all these organizations and there is no real differentiation among them. The same project can have multiple companies running it, with personnel being recycled from one to the other.
Because of this lack of differentiation, the only way for these companies to compete with each other is on the price of services. As they use the same people, the same managers, and the same approaches, the only way for them to compete is to cut corners. We hear about them in the news when they fail. Ethical and technological standards get lower and lower in this race downward.
One of these corners is the idea of minimizing the obligations under SOW's. Some of those SOW's promise little more than standard deliveries adding very little to what would be achieved by following the instructions in the installation manuals.
Often, to keep costs down, all personnel are haphazardly hired after the contract is awarded. The hiring managers have very minimal technical experience and technical interviews are more and more becoming a formality, reduced to buzz words and meaningless certificates.