The Future Revisited

By: Scott Bradner

The problem with making predictions is that someday one must look back and see what the batting average was. (To use a phrase a bit underutilized last year.) A year ago I tried my hand in the prediction business. (Competing with the Tabloids, Network World 1/10/94) It is time to see if reality and prediction had anything in common. Ranking from -3 (missed badly) to +3 (home run)

I predicted that the growth in Internet stories would exceed the growth of the net itself. Well, according to the St. Petersburg Times, the word "Internet" appeared 8,835 times in some collection of news sources during 1993 and 25,047 in the same collection in 1994. This is a rate of growth that is about twice the actual rate of Internet node growth. I get a +3 here.

I predicted that Apple would top its DCE refusal in trying to make its products more orthogonal to industry trends. Well, not so good here, Apple did over price their 7.5 system upgrade and did slip version 8 but the slip did match the rest of the industry. Apple did not seem to do much else positive or negative. I'll have to accept a -1 here.

I made a safe prediction in saying that the relevent vendors would continue to claim that their own protocol was better than TCP/IP and that TCP/IP would continue to outrun the competition for internetworking applications. A +3 for accuracy but moderated by the difficulty level. So a +1 for each.

My prediction of a major Internet security incident came all too true. (The pain of being right, Network World 2/21/94.) A sad +2 here.

Porn on the net was an issue but, knock on wood, only attracted CMU and not the 'crusading reporters' I predicted. A +1 here also.

I did include an ATM prediction, but since I was talking about ATM beginning to get cost effective, this was a long term prediction and can not yet be scored.

I also predicted that someone (anyone) in a telephone company would trip over a clue about how to offer Internet service. In my haste I did not make it clear that I was talking about the Bell-etts and not companies like Sprint who had already been offering such services. Well, the kids are still mostly trying to figure out which sand pile to play in while Sprint ,and now MCI, are well along in building sand castles. (Sand castles with concrete walls by the way.) So lost this guess, -2.

I predicted too, that we would have a rosy and wonderful world after Allison Mankin & I made our recommendation on IPng in July. Well, the recommendation was made and has now been approved by the IESG. The work of finishing the final details is proceeding on the actual standards and most of those will be ready in February. Things are proceeding quite well, not everyone agreed with 'our logic and powers of persuasion' but most have, and are now figuring out how to plan for an IPv6 future. I claim a +2 here.

I predicted that the annual router performance tests would come to an end because all the vendors would provide 'clear, believable performance numbers on their products.' Well, suffice it to say that the next round of testing starts in a week or so. -1 here

The final prediction last year was that finding topics for columns was going to get hard, +2 here.

So, lets see ... that works out to a +8 out of a possible +30 for last year. Not in Nostradamos' league but better than the horoscopes in the Boston Globe.

Oops, running out of space for predictions for next year. Quick, here is one before the end of the column. ATM will again fail to live up to its press billings this year. A sure +3 in '95.

Disclaimer: The above evaluation was not the result of any formal Harvard process, it is a case of the student grading his own exam.