I think one of the problems with DISMS is that it was conceived in the era just before the advent of personal computers, and consequently it was designed to be entirely a mainframe system. During this period when PCs were gradually entering the office space, the attitude of seasoned computer professionals towards them ranged anywhere from active curiosity to open contempt.

After my boss Allan Rosen abandoned me to take a more interesting job in another area, I became increasingly involved in DISMS projects, much against my will.

My exit strategy wasn’t working out too well. I had applied for a position in the Medical Directorate’s Management Support Office (MSO) as I thought I had a good shot there. They had previously selected me for a job that I applied for, but I turned it down because Colonel Lavin had gotten me a comparable position in C&T’s MSO (you know, the one where I ended up leaving on rather unpleasant terms), and I felt I owed it to him to give that a try. Anyway, I had made the list for this Medical position, but they were now taking their good old time in setting up the interviews. So it looked like I’d be stuck in DISMS for the foreseeable future.

So one day I was watching a demo of an upcoming DISMS release. As I recall this was for the editing of tech data, and the demo was being done by the DSAC people (that is, not the group of DISMS analysts that I worked with in Subsistence, but the programming group who dwelled in offices on a higher floor).

Just a reminder. The respective management teams (especially the military leaders) of the DISMS group and the DSAC group pretty much hated each other. This filtered down to some extent to the people who worked for each of them. Plus there were the usual bad feelings between the two groups working on a complex project that was continually falling behind schedule with everyone trying to blame everyone else. While some individuals were able to get along across the two groups, to a large extent there was little love lost between them.

Anyway I was watching this demo and the DSAC person seemed to be quite proud of this upcoming release where the tech data could be edited right there on the screen of the mainframe terminal in what had to be the clunkiest, awkwardest editor I had ever seen.

And I opened my big mouth.

Now stop the frame right there.

Over the years I have had a tendency to lack a certain amount of, shall we say, self control, when I get overly excited about something. And I’ll say the first thing that comes to mind without thinking it through. I think I’ve gotten better about it as time has moved on, but back in 1986 I clearly had not.

And so I opened my big mouth.

Now I’m not sure exactly what I said or exactly what happened during the rest of that meeting, but the upshot of it was I found myself tasked with coming up with a way to edit the tech data on a PC.

Did I mention that when Allan Rosen hired me, there was nary a word about getting involved in DISMS? I was just supposed to be the PC support person. Run the users’ meetings. Teach training courses. Solve PC related problems. Why did I have to go and open my big mouth?

So what to do?

I think I had probably outlined an approach in the meeting when I laughed at the demo. Instead of sitting at a mainframe terminal, the user would be working at a PC using terminal emulation software. She could either create the tech data in a PC text editor (if it was new data) or grab it from the screen and then edit it in the editor to make whatever changes needed to be made. Then she could press a key combination which would tell the PC’s software to take the text that was in the editor and inject it into the terminal software as if the user were typing it in herself.

I knew this kind of scripting approach could work because I used it at home with software to access CompuServe (an online computer service that predated AOL and the Internet). Almost right away I realized that to make something like this work, I’d have to implement only two commands: delete and add. To try to implement anything like an actual edit mode would raise way too many complications.

And so I brooded for several weeks, working with the terminal emulation software, tinkering here, tinkering there.

Finally, I went in to see the head of DISMS, a Navy Commander, whose name I’ve long since forgotten (sorry).

And I told him I couldn’t do it.

Oh, it could probably be done, and I could probably do it, but it would be a very fragile solution, subject to too many errors. Plus, any slight change that DSAC might make in a future release could very well break it, as might a future release of DOS or the terminal emulation software or who knows what. It just wasn’t a very smart way to go.

It was the one of the most humiliating moments of my life.

Later I remember talking to the woman from DSAC whose demo I had interrupted. She was surprisingly gracious at my failure.

Once again, what was needed was a way to exchange data with the mainframe and the PC, and I just didn’t see a way of doing that.

Yet.