SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Discuss Year 2000 Issues -- Ignore unavailable to you. Want to Upgrade?


To: John Mansfield who wrote (2065)6/28/1998 3:16:00 PM
From: John Mansfield  Read Replies (3) | Respond to of 9818
 
From Gary North' site - 'I am a Y2K consultant under contract with a fortune 50 company

'What Mr. Yannone Forgot to Mention....

'Comment:
This is from J. Ritchhart. jimr@starnetinc.com

* * * * * * * * *

I am a Y2K consultant under contract with a fortune 50 company doing Y2K work. I currently give speaking engagements and presentations on the year 2000 issue and have consulted with various different people concerning what they might do to prepare. I charge less that the current billing rate on my contract and I charge nothing for speaking engagements, presentations, and consults. I say this up front only to say that I am afforded some credibility in the consulting world working on micros and IBM midrange computer systems.

Mr. Yannone focuses on only one small issue of the total Y2K problem, that of various systems communicating with one another either compliant to non-compliant or different date formats across compliant systems. Even if I concede his argument, which I don't, it only addresses one issue (and a small one at that). Mr. Yannone does not address getting the job done in the first place. A thing many experts say will not happen. For example:

* Rep. Steven Horn said, talking about the federal government: "We are on the Edge of Failure"

* A study of 200 U.S. Businesses accounting for 10% of the GNP found that less that 1/2 are beyond the planning stage for fixing their Y2K problems and these companies rated themselves a 2.5 on a scale of 1 to 10 (10 being high Y2K readiness).

* IBM has said that the FAA computers will not work, cannot work, and cannot be fixed.

* As of 1st quarter of 1998, only 40% of the companies on the country even had a Y2K budget approved.

* According to the Office of Management and Budget, current estimates for 9 different federal departments has their Y2K repairs completed after 2002.

* According to Rick Cowles, an expert in the electrical industry, 56% of the electrical utilities won't even commit to a compliant date.

* Forbes mag. Reported that only 63% of the government's mission critical systems are expected to be completed on time and this was before Rep. Horn reported that the Gov. efforts had actually slowed down.

* Computer world reported on May 28th, that 50% of all businesses polled don't plan on doing anything about the Y2K issue. Presumably because of the cost and people like Mr. Yannone saying nothing will happen.

I could continue, but you get my point. Mr. Yannone's arguments could be mute if the above statistics are correct.

But what about Mr. Yannone's arguments? They seem to rely on one primary pre-supposition. That technology and programmers work in a perfect world. A world of validating incoming data, making sure every "I" is dotted and every "T" is crossed before processing any data into it's production environment. This pre-supposition makes since when you look at a recent survey done by information week. This survey measured the publics confidence in various institutions. Coming in last was insurance companies with 16 confidence points. But coming in first was the high tech industry with 750 confidence points, a good 250 points above the next highest category (research scientists). This might explain way I hear so many people say that "this is a technology problem, and technology will solve technologies problems".

But this is simply not the case. Even on my own contract, one of the systems that has been converted has approx. 25 systems that it passes data back and forth to. Very little validation of the data is done. Only a moderate amount of data validating is done before files are updated. Our system was completed and passed independent testing by an outside company. We were the first system to be certified compliant in the company. Brand new code that was tested and approved was moved into production on a different system and crashed my system because it passed us non-compliant dates. The next day, when I asked the other programmer how this date field was missed, she said "I was just so used to dealing with 2 digits years, I forgot". My point is that both systems were tested and approved by quality assurance. Both systems validate the data up to a point but not to the extent of validating the format of every field.

This is not an isolated case. In over 13 years of developing systems, the most validating that I have seen any system do is against master type information. This assures cohesiveness across files but not the format of the data in any one field.

The other major problem that Mr. Yannone demonstrates is his lack of objected facts and references. Everything listed above can be checked out. I have either listed specific references or sighted surveys done or quotes made. Mr. Yannone makes global and very general statements that have no objected facts backing it up. Statements like:

* "Everyone knows it's coming."

* "No company that has solved its own internal compliance problem is going to leave itself open to data corruption."

* "No programmer would subject his system to the vagaries of potentially corrupt data, so he would build in validation to assure that all incoming data is compliant."

These are general and vague, but he also contradicts himself by saying:

* Any data corruption from any source for any reason can disrupt a network of computer systems. This happens all the time. Programmers program for this eventuality routinely so that disruption is minimal and nonfatal. This is what error handling routines are for. They are in place already. It's standard practice. Programmers know this--historians have yet to learn about it.

As for me? Well, I am taking my money out of the bank, buying Gold, stocking up on food and doing other radical extremist type things. If nothing happens, well, I will have lots of food to donate to a missions organization oversees.

But if Mr. Yannone is wrong and does nothing. . . . .

See you on the other side of the millennium,

JR


garynorth.com



To: John Mansfield who wrote (2065)6/28/1998 3:42:00 PM
From: John Mansfield  Respond to of 9818
 
' Comprehensive List of Potential Y2K Problem Dates

Testing should include a number of critical dates to ensure compliance, so that no problems occur prior to, on, or after January 1, 2000. The algorithms of systems and chips need to be tested for forward and backward processing. Not all systems need to test for all dates listed. Different application domains may have specially significant dates like the fiscal year for business systems. It is up to the program managers to determine which are most likely to impact their systems. The most critical dates that should be considered for testing at this phase include:

November 2, 1997: Overflow HP/Apollo Domain OS

January 1, 1998: To ensure that the digits "98" do not trigger a red flag, result in erroneous branching, or otherwise cause a processing error or that "time error" faults occur. Also to ensure that December 31, 1997 was calculated as the 365th day of 1997. [Found in Y2K patches in mainframes and elsewhere.]

December 31, 1998 to January 1, 1999: To ensure that the digits "99" do not trigger a red flag, result in erroneous branching, or otherwise cause a processing error or that "time error" faults occur. Also to ensure that December 31, 1998 was calculated as the 365th day of 1998. [Found in Y2K patches in mainframes and elsewhere.]

January 1, 1999: Introduction of electronic version of the euro

February 5, 1999: 1st possible airline reservation problems (Max 330-day look-ahead)

Fiscal Year 2000 for Business and Industry: Depending on the business, the fiscal year could start on March 1, 1999, July 1, 1999 or match the government fiscal year of October 1, 1999.

August 22, 1999: Overflow of "end of week" rollovers (e.g., GPS). Unremediated GPS think it's 80/01/06.

September 9, 1999 (9/9/99 or Possibly 9999): To ensure the digits "99" or "9999" do not trigger a red flag, result in erroneous branching, or otherwise cause a processing error

September 30, 1999 to October 1, 1999: This is the last fiscal rollover prior to Y2K [for many including the US Government].

October 1, 1999: First day of fiscal year 2000 [for many including the US Government].

December 31, 1999: Last day before 2-digit year equals 00. Many systems will not operate correctly as they transition to the next day.

January 0, 2000: To ensure this date is NOT processed [some spreadsheets and database applications do have this problem and count January 0 as a day before the 1st]

January 1, 2000: Key date in any compliance testing

January 1, 2000, 1200 Hrs (Noon): Embedded date chip failure has been found

January 3, 2000: First full work day in the new year. First possible payday after rollover.

January 10, 2000: First 7 or 8 character date in YYY/M/DD format (2000/1/10 or 2000/01/10)

January 15, 2000: W2s due

February 28, 2000: To ensure the leap year is being properly accounted for. Many programmers have incorrectly been taught that the year 2000 is not a leap year -- Year 2000 IS a leap year. Systems should be tested to ensure correct handling of the transition to the 29th days of March 2000.

February 29, 2000: To ensure the leap year is being properly accounted for. Some systems may transition to the 29th of March 2000 correctly, but may not allow the date to be set to the 29th. This would happen if a system was reinitialized after the transition and should be explicitly tested for.

February 30, 2000: To ensure that this date is NOT processed [found in some PC applications]

February 31, 2000: To ensure that this date is NOT processed [found in some PC applications]

March 1, 2000: To ensure date calculations have taken leap year into account

April 2, 2000: First change to Daylight Savings Time after rollover (US)

April 15, 2000: Income Tax day

September 30, 2000 to October 1, 2000: This is the first fiscal rollover following Y2K [for many including the US Government].

October 10, 2000: First 8 character date using a 2-digit month (2000/10/10)

October 29, 2000: First return to Standard Time after rollover (US)

December 31, 2000: 366th day of the year 2000. This could be a problem for systems that use Julian dates.

January 1, 2001: First day in the 21st Century. This is the last leap year related date, testing the first day of January 2001 to ensure it can be set.

January 1, 2001: Overflow for Tandem systems

After January 1, 2002: Or any other date past this day, to ensure no processing errors occur in backward calculations and processing of dates in the 1980s and 1990s at this point in time

February 29, 2001: To ensure that this date is NOT processed as a leap year

February 29, 2004: To ensure that this date is processed as a leap year

January 1, 2010: Overflow ANSI C Library

September 30, 2034: Overflow of UNIX time function

January 1, 2037: Rollover date for NTP systems

January 19, 2038: Overflow of UNIX systems

September 18, 2042: Overflow of IBM System/360

2072, Exact Date TBD: Overflow of Milstar Operating System

February 28, 2100: Last Day of February - NOT a leap year

mitre.org