Quote:

"[Software remediation] is FATALLY FLAWED by its very nature no matter how competently done."

--Dale Way (IEEE Y2K Chair)

-- D.b. (dciinc@aol.com), November 17, 1999

Answers

quote on page 13 of 20, section 2.3, first paragraph

-- D.B. (dciinc@aol.com), November 17, 1999.

Make me

-- Rothog the Terrible (placeofearth@hog.cave), November 17, 1999.

...an instrument of your peace, Lord...

-- J (Jayho99@aol.com), November 17, 1999.

The guy is a philosopher more than anything. How can this statement (taken out of context, albeit) be true? Surely changes (remediation) can be made. The problem in the case of Y2K is that many, many remediation projects are being performed simultaneously and that not enough testing has been done. There will be many failures at the same time and this will contribute to anything from economic problems to societal breakdown. That's my understanding of the situation.

-- Mara (MaraWayne@aol.com), November 17, 1999.

Mara,

"TRADITIONAL SOFTWARE REMEDIATION BASED ON THE GOAL OF SYSTEM COMPLIANCE WILL NOT WORK in most cases, if the goal is protection from Y2K failures. It means NOT that remediation can fail because it can be done badly, which of course also happens, it means it is FATALLY FLAWED by its very nature no matter how competently done. Traditional software remediation is invasive, involving code modification of a working system. It is based on the Independence Fallacy of Y2K. By the actions taken and not the words said, it assumes the thing being remediated is an independent entity such that changes in its logic need only be considered as to their effect on the thing itself. This is the basis for much, if not most, in-house efforts, and all outside "Y2K factory" code scanning and repair services: "You send us code, we find dates and inject windowing code or modify code to account for data expansion and send it back to you." In neither case is knowledge of all other "systems" that share data, and therefore could be affected by these modifications, considered in the modification process."

From Dale Way essay.

Mara, A philospher!!?? I thought he was an Engineer

-- db. (dciinc@aol.com), November 17, 1999.

>>>> ** choke **<<<

-- I'm verklempt (talk@amongstyour.selves), November 17, 1999.

Mara,

Fatally Flawed...as a strategy, remediation had to be practiced by at least 85% of the the entire infrastructure. "Mission Critical systems are only 3% of the total, and we have not succeeded in remediating even that. I estimate that the average company is 3-6 months behind at this point. Look at the final 10-Qs at the SEC to get the unvarnished truth.

Glad to see that you moved from NYC...you are fortunate.

-- K. Stevens (kstevens@ It's ALL going away in January.com), November 17, 1999.

"it assumes the thing being remediated is an independent entity such that changes in its logic need only be considered as to their effect on the thing itself"

Sounds like a foolish statement to me. It doesn't correspond to any of the remediation I've seen, and I've been doing Y2k remediation since 1994, both manually and using automated search tools and automated code-fix tools. (The automated stuff is mostly worthless.)

No responsible or intelligent geek would think that fixing a system in isolation is the whole story. You always consider the outer boundary of its impact, and you clean up everything out to that point. You clean the program that writes a file, but also every program that reads that file. If windowing changes how a program will transmit data, you test the programs that receive it, etc. This is not tough to see, though IN PRACTICE hurried people might skip it.

Remediation is NOT flawed in its nature, it is merely executed poorly in many cases. But then, most system design is executed poorly, and the average program is written by a programmer who is, after all, only average.

-- bw (home@puget.sound), November 17, 1999.

I must agree with bw. If Way is correct, then the entire concept of code maintenance shares this same flaw. Which must come as a surprise to the army of code maintainers who've been doing their thing all these decades, with every appearance of total success.

Side effects of code changes are no mystery to programmers -- the very goal of almost all computer code is to produce side effects, such as performing calculations, doing IO, and generating output. Of course all these things can be done wrong. But the concept isn't wrong in principle.

-- Flint (flintc@mindspring.com), November 17, 1999.

Can you immagine discussing every sentence that every person ever wrote?

I mean, who cares and why is it important?

-- Cherri (sams@brigadoon.com), November 17, 1999.

Cherri,

Maybe Mr. Butt would prefer us not to be hoarding.

-- the Virginian (1@1.com), November 17, 1999.

Mr. Virginia,

Mr. Butts (thats Butts not butt) thinks youre not exactly the sharpest tool in the shed!!!

I think people should prepare to the max. where in gods name you construed my desire to discuss what is the most inflammatory essays about the debacle of y2k is beyond me.

Do yourself a favor---READ---threads before you respond. I cant tell you how weary I am of immature Adults and there funny little commments about the last name that I did not choose.

-- David Butts (dciinc@aol.com), November 17, 1999.

Butt butt butt butt....LOL sorry Dave, couldn't help myself.

-- Nikoli Krushev (doomsday@y2000.com), November 17, 1999.

it would seem that the severity, or lack thereof of Y2K effects might hinge on this one premise. - i.e.

he [Way] is either correct, and thus, so is Infomagic and we all take that course;

or he [Way] is wrong, and it's all been just a big waste of time.

not being competent to discourse, I won't.

again, however, I suggest a re-reading of Razon Robinsun's "Is underestimation a possibility" on the Humpty Dumpty board.

dynamite questions...

-- Perry Arnett (pjarnett@pdqnet.net), November 17, 1999.

Mr. Butts, I am always at the ready to apologize for an error and I truly did not leave out the s on purpose. Will you please forgive for my doing so?

-- the Virginian (1@1.com), November 17, 1999.

Butts are a favorite of mine .Don't worry David You could have a name like Buttsinski----it's not half as bad as my friend Richard Head.

-- H fats Kissinger (draconionsolutions@uselesseaters.com), November 17, 1999.

Talk about names designed to invite schoolyard abuse: When I was in school, two brothers from W. Germany moved into the neighborhood; the Hoff brothers; Dieter and JACK!

-- Ralph Kramden (And@AwayWeGo.com), November 17, 1999.

Hi bw and Flint. You're quite right in your assertions, this is a perfectly normal and well understood problem. But what's new about it is that the first real live test of many (not most, but many) of these remediated systems won't happen until 1/1/1900. It's a unique situation, but I'm viewing 1/1/1900 as being most like the start of field trials or beta test for these systems. That concerns me, lots.

-- Colin MacDonald (roborogerborg@yahoo.com), November 18, 1999.