Somebody went and asked them

Most industry history is written from documents: press releases, filings, product literature, and whatever the trade press printed at the time. All of it is real and none of it tells you why a decision was made, because that is not what any of those things are for.

Between 1988 and 1994, James L. Pelkey did something different. He recorded eighty-one interviews with founders, engineers, executives and regulators who had built computer communications, and he did it early enough that they were still working, still reachable, and still able to remember the reasoning rather than only the outcome.

The result is a hypertext history covering 1968 to 1988, freely readable, and a 2022 book edition written with Andrew L. Russell and Loring Robbins. The transcripts themselves are archived at the Computer History Museum as the James L. Pelkey collection — catalogue number 102746648, extent eighty-one items, credited as a gift from Pelkey.

Why the timing is the whole thing

Start that project in 2010 and most of those conversations are unavailable. Start it in 1988 and you get the person who made the call, describing the constraint they were under, before the story had been smoothed by two decades of retelling.

That window has closed. Whatever was not captured about the 1968–1988 period by the mid-1990s is now reconstruction rather than testimony, and reconstruction is what most of us are working from most of the time.

The idea worth borrowing, with his name on it

Pelkey's analytical contribution — separate from the interviews — is the market window: the observation that a technology succeeds or fails less on its merits than on whether it arrived while a market was open to it.

It explains something that a founding-year timeline cannot. Companies in this field cluster: a group appears around 1968, another around 1979, another around 1984. A list ordered by date shows the clusters and says nothing about why they are there. Pelkey's answer is that each window opens for a reason — a regulator's decision, an accumulated need, the success of the previous window — and closes when the question it formed around has been answered.

I have written three explainers applying that frame to the three windows, and his name is on each of them, because borrowing somebody's analytical apparatus and presenting it as your own reading of the evidence is not a small thing.

Where to start

If you have never read it, the useful entry points are the sections on the early interconnect decisions and on the 1988 trade shows, because both are places where the interview material says something the documentary record does not.

The work is at historyofcomputercommunications.info, and it costs nothing.