This appeared on Peter de Jager's forum, in response to David Hall's warning. I post it here because I believe we need as much information and expert opinion as we can get on this great unknown factor. One thing is sure: the programmers are not agreed.
* * * * * * *
From: Grayson Lynn Reply-To: Grayson@POBoxes.com To: email@example.com Subject: Re: Embedded Systems: Diatribe and Rant
Given that digital computers run on logic in its most elementary incarnation, one would reasonably expect that practitioners of the digital arts might think logically, or at least rationally, if only by osmosis... but alas not.
One camp asserts "there are needles in this [embedded systems] haystack" while the other maintains there are none. They wrangle on and on, for over a year now in this forum, without ever moving beyond that most fundamental of questions.
A single needle, a single live example of a millenium-impaired chip in a meaningful role, would settle that question and let the discussion move on to other questions... how many? how costly? how dangerous? where are the others?... but none is forthcoming.
As time goes on and the roster of candidate chips grows from millions to billions and beyond with still no winner in sight, it is hardly surprising that any observer with even the most primitive comprehension of probability would be led to believe that (1) nobody is seriously looking for that needle or (2) nobody would recognize a needle if they saw one or (3) nobody who has ever found a needle wants to admit it or (4) there really are no needles or (5) whatever needles there may be are rare and unimportant.
The status quo may be philosophically and psychologically entertaining, but it is scientifically absurd, and the executives and managers are entirely correct: show us a needle or get out of our face.