"can you PROVE beyond a shadow of a doubt the integrity of your database?"

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

Can you imagine a world where almost everyone (lawyers included) can say..."can you PROVE beyond a shadow of a doubt the integrity of your database""

-- thom (thomgill@eznet.net), January 08, 2000.

This is an excellent point from an earlier post today. This phrase will enter common parlance by the end of the year if/when database corruption becomes public knowledge, to be added to such malapropisms as:

"to my knowledge..." [well, whose other knowledge would you have?]

"it depends on what you mean by "is"...["is" that having sex?]

-- Total Doomer (sky@falling.com), January 08, 2000


The bad data is churning and turning and being passed on. We will slowly grind to a crawl, we may be experiencing devolution already but not realize it.

-- Porky (Porky@in.cellblockD), January 08, 2000.

What's amazing is that, truly, a tree did NOT fall until an accountant has to account for the loss of shade, or the unusual accumulation of leaves on her desk.

Trust me: I've always been the chick that fixes accounting problems... usually they don't start stomping into your office until the problem is at least a month old.

Three weeks of respite.

-- lisa (lisa@work.now), January 08, 2000.

"Can you PROVE..."

Interesting that this little collection of words is what got the Y2K doomers into trouble..."Can you PROVE...Y2K will not cause major disruptions..."

The truth is, no. I can't PROVE I won't be killed by a stray bullet in the next five minutes etc, etc.

What I can reasonably decide, though, is that a computer failure won't crash the world.

-- ImSo (happy@prepped.com), January 08, 2000.

IMSO, it will crash the world soon,the dependence from the machines is getting every day larger and larger we will becrashing if we do not fix this bug,that crawls like a fever tha killed milipons in Europe,like Aids,like I do not know what We must Kill The beast!!!

-- Alex The Italian (alexcapr@yours.com), January 09, 2000.

What I can reasonably decide, though, is that a computer failure won't crash the world.

Hope there's no critical mass for enterprise system errors. I, for one, could not operate a two-million dollar business based on hand-written checks by high school dropouts, functional illiterates and (ex)-cons.

-- Total Doomer (sky@falling.com), January 09, 2000.

(con't): My business is totally dependent on cash flow from enterprise-scale systems. Back-up employee cash payroll will remain undeposited at least until the end of the month. I suggested to not keep cash reserves for my DGI employees or spouses, but my business manager said that was unfair.

-- Total Doomer (sky@falling.com), January 09, 2000.

Reply to ImSo's comments on this thread.

It is obviously impossible to "prove" a future event will occur when there is only a chance...even if that chance is unknown.

Regarding being hit by a bullet...I was told that the Japanese Ki Master who founded the Aikido movement was able to stand in front of expert marksmen and let them shoot at him...and he dodged the bullets.

The world I was discussing in this little "quote" attributed to me was a world where things have already happened.

Like you owe $1000 more on your credit card...and you are disputing charges....the company says you owe...you say you don't. Your bank seems to have lost all record of your mortgage history and to boot has replaced your loan with the figures from a different loan.

Until now, the guy with the computer has been able to say "well it is in our computer"...and you might say "do you have a hard copy?" and usually the hard copy is a printout of the same database. Now, the guy shows you the printout and you can say "Can you prove the integrity of your database?"

How about a county which has rolled back its computers to 1999 so they will accurately manage data? An employee calls to ask about their check for March which was not issued. And the clerk says..."sure it was issued it says right here check for March 2000 issued on Dec. 31, 1999." Who has the apparent moral authority?

Let's take one example...a felon who ha

-- thom (thomgill@eznet.net), January 09, 2000.

"Regarding being hit by a bullet...I was told that the Japanese Ki Master who founded the Aikido movement was able to stand in front of expert marksmen and let them shoot at him...and he dodged the bullets."

When the Ki Master is ready to show you how powerful he is by breaking a piece of wood with his hand, walk up to the boards on the platform and turn them around so the end grain of the woood is resting on the support. Bet it doesn't break. Anyone with a bit of leverage and knowledge of wood can break a board with the grain.

There was a bar room show man (saw this on TV) who was impressing everyone by catching bullets in his teeth. Finally a drunk patron pulled his gun out and said, "Catch this one." Only to have another tip his gun to shoot in the ceiling. The show man left and never returned.

Another Ki type guy showed us how he can bend a sword with his belly. Take a real tempered sword up to him and press it to his belly and ask him to push. If the sword snaps in two then I will be impressed.

In other words, "Don't believe everything you see and hear, especially on tv."

-- bull [et] poop (bite@the.bullet), January 09, 2000.

First let me say that data integrity is important. However, data corruption comes in many degrees and it's not enough to ask "Is the data corrupted?" for the answer for a great many databases in use today is yes, it is corrupted. Always has been. Some data is much more important than other data (e.g. financial transactions are more important than advertising mailing lists).

Where I work, there are 2 people employeed pretty much full time in the task of data hygiene. That is, their job is to continuously work with the various data bases that we administer and to remove known errors.

As nearly any database administrator will tell you, almost no database is up to date and free from errors. From a practical standpoint, a perfect database is not required for many tasks (I'm excluding financial transactions).

So again, this is not a binary issue. It is more an issue of degree of corruption and, more importantly, the impact of the corruption. Frequently, data corruption leads merely to higher than necessary costs and nothing worse. Of course, this may not be of much comfort if you are the one suffering the increased costs.

We (globally) already tolerate a significant degree of data corruption. Whatever, problems come our way as a result of data corruption, we are likely to be able to respond in a reasonable fashion. Delays in some things? I suppose that's likely but I don't see it being anything we can't handle with a little patience, preparedness and understanding.

-- Arnie Rimmer (Arnie_Rimmer@usa.net), January 09, 2000.

Proving the integrity of a database even without Y2K is a messy business. That's the kind of stuff CS people get to think about if they do something stupid like enter graduate school.

-- Tim the Evil Programmer (tmiley@yakko.cs.wmich.edu), January 10, 2000.

Moderation questions? read the FAQ