A message close to home

SMTP is pretty much a house-hold term.  My wife, a non-technical person, was asking me questions on SMTP when trying to connect her mail agent on her smart phone to her Google email-account.  Referring back to Wikipedia, SMTP (Simple Mail Transfer Protocol) can be traced to 1960s.  Communication between mainframe computers used electronic messaging.  As more computers were interconnected, especially in the US Government’s ARPANET, standards were developed to allow users of different systems to e-mail one another. SMTP grew out of these standards developed during the 1970s.  Ray Tomlinson of BBN invented for TENEX computers to send mail messages across the ARPANET.  I actually had the privilege of using a mini-computer at Annenberg School of Communications, USC in 1977 to send a message to a friend in Indian Institute of Technology, Madras.  Development work continued throughout the 1970s, until the ARPANET converted into the modern Internet around 1980.

 Let us forget about my attempts to commercialize it, but what I was not aware that Jon Postel published the specifications for SMTP in RFC 788 in 1980, and later in 1982, RFC 821.  Jon Postel was part of the Information Sciences Institute, University of Southern California.

Amazing that over three decades ago, I had touched a piece of technology that would become one of the dominant pieces of collaboration today.  Proud to be a Trojan!

A massive IT Project Failure?

The blame game in project failures continues.  Historically, on larger IT projects (exceeding $15 million) about half are over budget and roughly half deliver less functionality.  If you look at the failure of the $100 Million student data project funded by the Gates Foundation, there are many lessons to be learned.   The purpose was noble – use student individualized educational data to streamline the education process.  The reasons given are that it failed because of politics and not related to the project approach.

The most obvious one, not mentioned in the analysis is the lack of “experimental innovation” i.e. small connected projects versus one big data collection approach chosen by inBloom, the company funded to implement the data collection project. A simpler, less costly and easier to adopt approach would have been to use existing well defined tests (propagated through contests) to identify individualized education that targets each student.  This was the approach we followed with great success at the Braille Institute on a Federal grant significantly smaller than the Gates funded research.

A second reason is failing to take into account the challenges of change management.  Tools are an enabler of change, but change management can only be effective if the user needs are well understood and identified early in the project.

There are perhaps many more reasons for the project’s failure such size and scope being too large, but the lack of small incremental projects seems to be strongest reason for the failure of this large project.