Tuesday, November 30, 2010

Oops types

So, we need to address this issue so that we can motivate solutions. That is, yes, the tone here will turn from taking potshots at oops and the oops'ers toward a more constructive bent.

We can talk three basic oops types, for now. There will be more which then will be categorized, analyzed, and smoothed (nod to the fact that underdetermination is the main thing, folks) out (we cannot remove wrinkles of many sorts).

So, for the first type, it would be a purely human affair, like the hubris we see with the CEO crowd (not all, just those with the biggest egos and pockets). But, there are other types. Is not the saying, to err is to be human? Does that not imply that none of us are without some fault and mangling ways?

In regard to that, the CEOs ask for their comeuppance several ways which we can discuss at length (but will not, given the new spirit). Trash talking gamers might be another category, but that brings in a whole set of other issues.

So, we have human error.

Then, for the second type, we have artifact oops, such as computer, or system, error. Mind you, system would imply products of all types. So, engineering has been working for eons on improvements. And, we see processes in place to help ensure better lives from use of the results. Some of this is just plain common sense, yet a lot of it had to be mandated by government actions (thanks, CEOs - oops, there I go again -- the new spirit will eventually win out here, it is guaranteed - oops, again).

Many product makers think about safety and minimizing harm from failure. Those who think forwardly in computing are looking at what would be involved with error-correcting systems. Mind you, communications has worked on this for years, with various types of checks and fix ups. But, as we ascend the conceptual tree (and complicate contexts), it becomes more of a problem, especially since undecidability is at the core. There have been means proposed, discussed, and implemented throughout the years. In fact, default logic, for one, was motivated by these concerns.

So, those are two types of oops: the wetware type and the hardware type.

What then is the third type? Well, sci-fi aficionados will get a metaphor of the borg. What we have, folks, is that ubiquitous computing has permeated human affairs to the deepest level and across the board. And, the growth proceeds both in depth and in breadth. There is no sense in being alarmed, as this progress was an inevitable consequence of the enlightenment switch that was thrown a few centuries ago.

It is that third type of oops that we'll focus on in 7oops7. Plenty harp on what others do wrongly, that seems to be a human trait. There is a saying about this that involves glass houses (find it, if you must). Many suffer from hardware (all types) failures. What is problematic there is that many providers of these things like to blame oops type 1; it's called user error.

Now, to be complete, we need to mention software. You see, with computation, even hardware can be parameterized. So, we will place software oops into the type 2, as then oops type 2 would actually be any type of system error.

So, we will look at oops of this third type, yet that we can differentiate between oops type 2 and oops type 3 is not always easy (here we need to link to discussion of cyber-physical issues). Come to think of it, what we might think of as oops type 1 could very well have some oops type 3 in it (that is, someone (CEO) takes a stance based upon they're told by a techie who is off the mark).

The main thing to remember is that oops type 3 deals with humans dealing with systems. This can range from comic (Chaplin's guy) to tragic (too many, and too depressing, to itemize). These types are going to become more of a concern (already were, but now cannot be ignored) as we proceed with the 21st century.

Aside 1: That examples of oops type 3 will be heavily computationally flavored is only the result of the growing ubiquity. In many cases, we might discuss things that are mechanical or other.

Aside 2: it might be fun to look at what the human-machine interface people have defined over the years in terms of oops, such as usability.

Remarks:

01/15/2012 -- Two oops are of interest here. One deal with a product, namely one of these types handled by Sealy. Now, another was related to computing, namely mad-house rush, on the web, to be first in line to receive something (Dr Oz). We saw mayhem this past fall in situations where people had to fight to get one of a limited number of goods (black friday). So, is that the point, the settos? If so, tsk?

02/03/2011 -- This is a place holder, for now, for Lewis' article. The Irish people (where is the rage?) were screwed over even more than the Americans. Now, one could argue oops; but, the truth is that certain minds need much more restraint than they are willing to admit. Unfortunately, other people bear the effects of these idiots (who, by the way, may, in many cases, test well - too bad there is not an effective arse test).

12/14/2010 -- One necessary topic will be compiled knowledge.

Modified: 01/15/2012

No comments: