Systems

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

Folks:

We are dealing with a systems problem. This is a complicated problem, but not the most complicated. Dealing with it is difficult because of the lack of facts. But it is not the most complicated systems problem.

Others include the climate, rain forest ecology, marine ecology, and, specifically, what is happening to the salmon.

For the latter, I can give you a lot of facts [but I won't]. Where there are dams, there are fewer salmon. Easy choice. So what happened to the 1988 sockeye run in Bristol Bay [no dams] or the Chinook run in the Yukon [same answer]. See, we are dealing with the system, not the runs on the Snake river. We are dealing with the whole system. It not only includes the breeding sites but what happens in the whole life cycle of the salmon [Cherri and Flint are without a clue in this area].

That is what we are talking about here. No matter how good you are at writing code, you are not good enough to determine what will happen [beginning at UTC turnover]. It is a systems problem. It is all a crap-shoot.

Best wishes,,,

G

-- Z1X4Y7 (Z1X4Y7@aol.com), December 04, 1999

Answers

Folks:

Evidently, so am I. That should read 1998 Bristol Bay run. The point is the Dams on the rivers have an effect, but they are most likely only one piece of the clue. Just like changing code in one computer. The system rules. If the system has many faults, changing one won't solve the problem.

Best wishes,,,

-- Z1X4Y7 (Z1X4Y7@aol.com), December 04, 1999.


Let me add to what Z is saying with my golf swing analogy. A golf swing is fairly simple. The ball doesn't move, the golfer is stationary, there are few moving parts. Now, when I was beginning the torture I used to spend a lot of time at the practice range, but I wasn't good at practicing because I often altered more than one part of my game at the same time. I might move my stance forward or back, change the alignment of my feet so that my stance was more open or closed, I could change my grip so that it was stronger or weaker. Any one of those changes would have an effect upon my swing, for better or for worse. When I made those changes together there was no way that I could keep track of the effects and my golf game was laughable. If the optimists are correct, then the whole global system built with and reliant upon computers will be able to withstand the changes of some sub-systems working properly, some malfunctioning immediately, some malfunctioning over time, some malfunctions being contained locally, some transmitting themselves to other systems in obvious ways, some transmitting themselves to other systems in hidden ways. Me, I have started unintentionally hooking the ball after almost thirty years of fading it and I don't think my swing is harder to remediate than Y2K. Cheers, AGF

-- drac (greenspanisgod@frb.giov), December 04, 1999.

Drac:

Good point. I hope that your golf game is better than mine. I can hook and slice at the same time. I am at once an embarressment and a wonder to my teacher.

Best wishes,,,,

Z

-- Z1X4Y7 (Z1X4Y7@aol.com), December 04, 1999.


Folks:

It'd not A system problem! It's MILLIONS of systems problems, with a common cause: DATE stamping, date formatting, processing of dates and date increments. Therefore, it systemic.

Since this problem affects "systems" in different ways, a huge effort of fixing (and testing!)is (was) required for ALL the systems to work right.

That's why death by a thousand cuts it a possibility if we rely on fix-on-failure, and that's why silver bullets can't work.

I do not accept that it is difficult due to "lack of facts" as posted above. In salmon migration, global warming, and such, the problem is the LACK of KNOWLEDGE of ALL Influential Parameters.

In the technical world, systems designers claim to have that knowledge of their systems (if they don't, they usually pay for it, unless they make a few quick moves.)

If you can't have facts for the future (absurd anyway), look at precedence! The designed systems aren't perfect, failures occur and are NOT ANALYZED by the media (if they are spectecular, the media may "report" their understanding of them.) Systems may not have been tested sufficiently for the inevitable parameter(!) CHANGE. Now they WILL -ALL(!)- be tested for their susceptibility to that parameter (TIME handling.) In this context Systems and Processes are the same.

Never before were all Systems and Processes subjected to a change at the same time! Never again will so many of them require immediate fixing. The inability to do so in three days is what's ahead. Never mind what was deemed mission non-critical before it failed. What will get us:

Lack of documentation. Lack of qualified people to do the job(s!) Lack of replacement parts. The neccessity for contination of business while things are being fixed. The most interesting: possible lack of reliable power while fixing IT systems!

All this in combination looks ideed AS THE MOST COMPLEX job we have ever encountered!

-- W (me@home.now), December 04, 1999.


The main difference between y2k and salmon runs is their level of liquidity or evolveability. A salmon run is evolutionary; it is liquid. Eventually, the salmon will re-estabish themselves in the rivers after we are all gone. Biology easily morphs into uninhabitied niches. On the other hand, technology is aliquid; it cannot evolve without the help of us. It isn't self-evolvable, because it is highly structured and programmed. Unless humans have installed high degree of back-up systems and simple work-arounds, the system is aliquid and therefore very difficult to patch-up a system-wide failure.

-- hindsight is 20/20 (76651.3647@compuserve.com), December 04, 1999.


Its not a systems problem, and its not a management problem. Its a human nature problem.

-- a (a@a.a), December 05, 1999.

Hindsite:

If there are no more salmon, they will not re-establish. That is true in many rivers. Actually, the problem is greater there. We can recreate computer systems.

W:

Can't agree. Don't know the difference between many systems and a systems problem. Your argument is a matter of decreasing analysis to the smallest unit. You don't seem to understand that this is the basis of the problem. I do value the input from programmers. It is just, that their knowledge is, in many cases, limited to the smallest subunit or a class of subunits. They, nor I, have the knowledge to deal with the big system [or system of systems if you prefer].

Best wishes

-- Z1X4Y7 (Z1X4Y7@aol.com), December 05, 1999.


Z1, my comments:

"You don't seem to understand that this is the basis of the problem." [YOUR interpretation of what I understand! Not so.]

"I do value the input from programmers. It is just, that their knowledge is, in many cases, limited to the smallest subunit or a class of subunits. They, nor I, have the knowledge to deal with the big system [or system of systems if you prefer]."

Right! Neither do the "programmers" that screw around with nature's "processes" realize that they only look at a very small increment in time.

The difference is that technical processes were designed from scratch by humans, but nature's events were not. Looking at them is interpretation of the whole by observation of an increment.

The point is that what WE created was based on knowledge, and what God created is still being anayzed to gain knowledge.

I think that by now, we know pretty well what is going to happen to data processing if the dates are NOT handled correctly. We just don't know what the secondary effects will be. But the educated guess is that productivity WILL get hit, and that in January, we will start all over with assessment, meetings, and plans to fix it. Same structure: top-down, guy in trench is told what to do. Guy in trench looks at sum of smallest increments, guy at top is a bottom-liner. Deadlines are set by the top. Essentially, we are going to use the same structure of command to do what we missed doing right before.

Suddenly, we have the knowledge of big sytems that we didn't have before the date change(?). If the programmers see only their tiny increment, and the top sees that the sytems need fixing, who is in the middle to do that? (I LOVE the question: "what are your qualifications", because it is so exluding.) As they blame each other for not understanding (see above), the HUMAN side of the event will play out fully. That is scary.

-- W (me@home.now), December 06, 1999.


Moderation questions? read the FAQ