Process Programming Deja Vu

The ICSE 97 program committee jointly honored Professors Leon J. Osterweil of the University of Massachusetts, Amherst, and M.M. Lehman of the Imperial College of Science and Technology, England, as having written the most influential papers for ICSE 9. Osterweil's paper and presentation 10 years ago, followed by Lehman's contrasting remarks spurned a debate the split the attendees at the conference that year in Monterey California. As Osterweil stated in his retrospective presentation, he was glad to see that the rock he cast had created a ripple that lasted the test of time.

1. I was pleased to hear that I was to receive the recognition. I do feel that the paper had a lot of impact, so this was not a total surprise. But it is always nice to hear that the community recognizes one's work.

2. and 3. I have vivid recollections of ICSE 9 and the controversy that was kicked up by the talk, and Manny's response. On one hand it was a bit jarring to be thrust into the middle of an impassioned debate. But on the other hand it was nice that the talk got serious attention. I thought that the ideas were valuable, but that is always what one thinks about one's talks. Sometimes one gives a talk and nobody seems to notice or care, and that is a disappointment. In this case, however, it was gratifying to hear so many people talking about the talk. I was surprised at how much the ideas seemed to be misunderstood and mistrusted. I remember thinking that I should have done a better job of explaining them. I would have liked it better if more people had lined up in support of my ideas, but I did have the distinct feeling that in time people would come to understand the ideas better and accept them. I like to think that that is what has happened.

4. In the past ten years I have gained a lot of insight into the original suggestions and proposals. That insight, coupled with ten years' experience in reexplaining what I have had in mind, seems to have put me in a position to do a far better job of explaining myself. I hope that is what I have done in the current paper, and in the talk that I will give on Wednesday.

5. As my paper has indicated, and as my talk will emphasize, I think the software engineering research community has acquitted itself well in debating with passion the ideas that were put forward. There are some deep ramifications arising from what has been suggested and they need to be debated and experimented with. The community continues to do that, and in doing so seems to me to be acting in the most important tradition of serious scientific inquiry. I think it attests to the continuing coming of age of software engineering as a maturing scientific discipline.

No time for a very careful revision of what I wrote to you earlier. But there is one additional thought.

This is probably an addition to answer 4 (I think), namely what might I have said earlier that I didn't:

I reread the original ICSE9 paper and was struck by the emphasis that I put on detailed encoding of the process in a coding language. In my concern to emphasize the need for something beyond designs, models, etc. of the process I leaned a bit too heavily on the importance of complementing such models with details that can be gotten from a code representation. In retrospect this left the suggestion open to the criticism that what was suggested waws the use of programming languages in place of models. In fact the title of the paper, "Software Processes are SOFTWARE too" was intended to strongly project the impression that processes are all that other software are--namely requirements, designs, models, code, etc. The paper did not carry this through as strongly as it should have. A lot of misunderstanding followed. I hope that the paper and presentation at ICSE 97 will help to set things straight.