The Institution of Electrical Engineers re Y2K: The Millennium Problem in Embedded Systems 4-2-2001 thru 2100

greenspun.com : LUSENET : Grassroots Information Coordination Center (GICC) : One Thread

The following is dated 4-2-2001.

http://www.iee.org.uk/PAB/SCS/problemdates.htm

The Institution of Electrical Engineers

The Millennium Problem in Embedded Systems

Dates Potentially Causing Problems in Computer Systems (from today to 2100)

Published 2nd April 2001

2001/09/09: (At 01:46:40 GMT) A UNIX date when the time_t value goes from 9 to 10 digits. Suspects are timestamps stored in fixed-column tables and internal variables. 2002/01/01:

(or any other date past this day) Processing errors may occur in backward calculations and processing of dates in the 1980s and 1990s at this point in time 2002/06/30: Last day European national currencies are acceptable. 2002/07/01: First day of Euro-only transactions in the European Union (EMU). 2004/02/29: Last day of February - a leap year. 2004/07/17-18: GPS Receiver Almanac Rollover due to the use of an 8-bit field for weeks since 22 August 1999 (256 weeks). 2004 & 2005: Children born in 2000 enter school system. 2005: Reported that some very old versions of UNIX (e.g. 16-bit BSD) will die in 2005. 2009/01/01: Satellite processing of distress signals from 121.5/243 MHz emergency beacons ends. 406 MHz will be used instead. 2009/09/09: Possible valid nonsense or marker date (same for any date where Y, M, and D are the same). 2010/01/01: Overflow for ANSI C library. 2010/01/01: Sorting YYMMDD decade-reversed covers 1990-2009 only. 2017: Children born in 2000 will be processed by university admission systems. If they are still in operation, systems may think they are dealing with 1900 or 2100. 2019/01/01: Possible confusion may occur between YY and YYYY forms, over first two digits of the date. 2019/12/31: YY date limit for Microsoft Excel 95. 2020/01/01: Mac (System 6.0.4+) Date and Time control will no longer be able to set the current date. 2025/12/31: Dates will fail on versions of Intuit's QuickBooks for DOS. 2027/12/31: Dates will fail on versions of Intuit's Quicken, QuickPay 3 for Windows, and QuickBooks for Windows and Mac. 2028/01/01: Systems that used a 28 year setback to remedy Y2K will fail. 2030: As with 1970, this has been reported as a breakpoint in the windowing system used by Microsoft in a large number of their products. In these systems 29 will imply 2029 and 30 will imply 1930. 2034/09/30: Overflow for UNIX time function. 2035/12/31: Microsoft's Year 2000 statement of compliance timeframe ends at 24:00. 2036/02/06: (06:28:16 GMT) 2**32 secs from 1900/01/01 - 00:00:00 GMT. NTP timestamp overflows (2036/02/05 if 1900/02/29 is taken into account). 2037/01/01: Rollover date for NTP systems. 2037/12/31: Year 2000 support for some editions of WS_FTP ends. 2038/01/19: (03:14:07 GMT) 2**31 seconds from 00:00:00 GMT, Thursday 1st January 1970 (UNIX's birthday). The seconds counter used for date/time information in UNIX and C and C++ will reach 2,147,483,647 - the largest number which can be stored as a 32-bit signed integer. As a result an overflow problem will occur (i.e. the value of the next number is unpredictable). Time differences mean it appears that it will happen earlier in America (22:14:07 US EST, Monday 18th January 2038). There will also be a an additional problem due to a discrepancy of a few seconds between system clocks and astronomical time because the system will not have taken account of leap second adjustments made in the interim period. This may be significant only for a few very specialised systems, but could give rise to difficulties if changes are based on algorithms which do not take this into account. It is understood that the Java programming language (which in many respects closely resembles C++) will not have this problem. 2041/01/01: IBM mainframe internal clocks will not go past this year (unconfirmed). 2041/11/16: (24:00:00) Clock in Unisys BTOS/CTOS Operating System will move on to 1952/03/01. 2042/09/18: (23:53:47 GMT approx) Overflow of TOD timer on IBM Systems 370 and 390. 2046/01/01: Amiga system date failure. 2048/01/19: 2**31 seconds from 1980/01/01 - Stratus VOS OS failure. 2048/07/01: 64**2 weeks from 1970 - some UNIX password ageing fails. 2049/12/31: Microsoft Project 95 (and earlier) limit. 2050/01/01: YY-windowing into 1950-2049 collapses. 2060/01/01: The trick of using a two-digit year representation with the first digit hex (98, 99, A0, A1 through to F9) fails today. If the digits are stored as nibbles, no more can be done; if as characters, 200 more years brings the end of Z9. 2068/01/01: YYMMDD = 000101 again for SunSPARC: SunOS, Solaris, BSD/OS, Linux: in RTC. 2070/01/01: Centenary of 1970/01/01 (base date for UNIX time_t). There is the possibility of Y2K style faults. 2071/05/10: (11:56:53.684) AS/400 internal hardware clock rolls over to 1928/08/23. 2072: (Exact date t.b.d.) Overflow of Milstar Operating System. 2078/12/31: Excel 7.0 - The Last Day, #65380; and Excel 95. 2079/06/06: 2**16 days from 1900/01/01. 2080/01/01: MSDOS file dates become ambiguous when displayed with two digit years. Windows File Manager, set to ISO-8601 dates, drops 100 years from displayed file dates beyond 2080. 2100/01/01: Y2.1K - most current PC BIOS run out of dates. MSDOS DIR renders filedate years 2100-2107 as 99. Many short-term Y2K fixes will fail. 2100/02/28: Last Day of February - NOT a leap year. First failure of the "4-year" rule since 1900 [NB - every year divisible by four is a leap year unless it is divisible by 100, but not 400]. 2101/03/01: It is claimed that a system which is compliant up to this date will be totally and permanently date compliant. [Main 2000risk page]

Document last updated: 2001-04-02 Document valid until: 2002-04-02

Page created by phawes@iee.org.uk

End of exerpted webpage



-- Paula Gordon (pgordon@erols.com), May 02, 2001

Answers

A retry to preserve the formatting:

Subject: The Institution of Electrical Engineers re Y2K: The Millennium Problem in Embedded Systems 4-2-2001 thru 2100

The following is dated 4-2-2001.

http://www.iee.org.uk/PAB/SCS/problemdates.htm

The Institution of Electrical Engineers

The Millennium Problem in Embedded Systems

Dates Potentially Causing Problems in Computer Systems (from today to 2100)

Published 2nd April 2001

2001/09/09:

(At 01:46:40 GMT) A UNIX date when the time_t value goes from 9 to 10 digits. Suspects are timestamps stored in fixed-column tables and internal variables.

2002/01/01:

(or any other date past this day) Processing errors may occur in backward calculations and processing of dates in the 1980s and 1990s at this point in time

2002/06/30:

Last day European national currencies are acceptable. 2002/07/01: First day of Euro-only transactions in the European Union (EMU).

2004/02/29:

Last day of February - a leap year.

2004/07/17-18:

GPS Receiver Almanac Rollover due to the use of an 8-bit field for weeks since 22 August 1999 (256 weeks).

2004 & 2005:

Children born in 2000 enter school system.

2005:

Reported that some very old versions of UNIX (e.g. 16-bit BSD) will die in 2005.

2009/01/01:

Satellite processing of distress signals from 121.5/243 MHz emergency beacons ends. 406 MHz will be used instead. 2009/09/09: Possible valid nonsense or marker date (same for any date where Y, M, and D are the same).

2010/01/01:

Overflow for ANSI C library.

2010/01/01: Sorting YYMMDD decade-reversed covers 1990-2009 only.

2017:

Children born in 2000 will be processed by university admission systems. If they are still in operation, systems may think they are dealing with 1900 or 2100.

2019/01/01:

Possible confusion may occur between YY and YYYY forms, over first two digits of the date.

2019/12/31:

YY date limit for Microsoft Excel 95. 2020/01/01: Mac (System 6.0.4+) Date and Time control will no longer be able to set the current date.

2025/12/31:

Dates will fail on versions of Intuit's QuickBooks for DOS.

2027/12/31:

Dates will fail on versions of Intuit's Quicken, QuickPay 3 for Windows, and QuickBooks for Windows and Mac.

2028/01/01:

Systems that used a 28 year setback to remedy Y2K will fail.

2030:

As with 1970, this has been reported as a breakpoint in the windowing system used by Microsoft in a large number of their products. In these systems 29 will imply 2029 and 30 will imply 1930.

2034/09/30:

Overflow for UNIX time function.

2035/12/31:

Microsoft's Year 2000 statement of compliance timeframe ends at 24:00.

2036/02/06:

(06:28:16 GMT) 2**32 secs from 1900/01/01 - 00:00:00 GMT. NTP timestamp overflows (2036/02/05 if 1900/02/29 is taken into account).

2037/01/01:

Rollover date for NTP systems.

2037/12/31: Year 2000 support for some editions of WS_FTP ends.

2038/01/19: (03:14:07 GMT) 2**31 seconds from 00:00:00 GMT, Thursday 1st January 1970 (UNIX's birthday). The seconds counter used for date/time information in UNIX and C and C++ will reach 2,147,483,647 - the largest number which can be stored as a 32-bit signed integer. As a result an overflow problem will occur (i.e. the value of the next number is unpredictable). Time differences mean it appears that it will happen earlier in America (22:14:07 US EST, Monday 18th January 2038). There will also be a an additional problem due to a discrepancy of a few seconds between system clocks and astronomical time because the system will not have taken account of leap second adjustments made in the interim period. This may be significant only for a few very specialised systems, but could give rise to difficulties if changes are based on algorithms which do not take this into account. It is understood that the Java programming language (which in many respects closely resembles C++) will not have this problem.

2041/01/01: IBM mainframe internal clocks will not go past this year (unconfirmed).

2041/11/16: (24:00:00) Clock in Unisys BTOS/CTOS Operating System will move on to 1952/03/01.

2042/09/18: (23:53:47 GMT approx) Overflow of TOD timer on IBM Systems 370 and 390.

2046/01/01:

Amiga system date failure.

2048/01/19:

2**31 seconds from 1980/01/01 - Stratus VOS OS failure.

2048/07/01:

64**2 weeks from 1970 - some UNIX password ageing fails.

2049/12/31:

Microsoft Project 95 (and earlier) limit.

2050/01/01:

YY-windowing into 1950-2049 collapses.

2060/01/01:

The trick of using a two-digit year representation with the first digit hex (98, 99, A0, A1 through to F9) fails today. If the digits are stored as nibbles, no more can be done; if as characters, 200 more years brings the end of Z9.

2068/01/01:

YYMMDD = 000101 again for SunSPARC: SunOS, Solaris, BSD/OS, Linux: in RTC.

2070/01/01:

Centenary of 1970/01/01 (base date for UNIX time_t). There is the possibility of Y2K style faults.

2071/05/10:

(11:56:53.684) AS/400 internal hardware clock rolls over to 1928/08/23.

2072:

(Exact date t.b.d.) Overflow of Milstar Operating System. 2078/12/31: Excel 7.0 - The Last Day, #65380; and Excel 95.

2079/06/06:

2**16 days from 1900/01/01.

2080/01/01:

MSDOS file dates become ambiguous when displayed with two digit years. Windows File Manager, set to ISO-8601 dates, drops 100 years from displayed file dates beyond 2080.

2100/01/01:

Y2.1K - most current PC BIOS run out of dates. MSDOS DIR renders filedate years 2100-2107 as 99. Many short-term Y2K fixes will fail.

2100/02/28:

Last Day of February - NOT a leap year. First failure of the "4-year" rule since 1900 [NB - every year divisible by four is a leap year unless it is divisible by 100, but not 400].

2101/03/01:

It is claimed that a system which is compliant up to this date will be totally and permanently date compliant. [Main 2000risk page]

Document last updated: 2001-04-02 Document valid until: 2002-04-02

Page created by phawes@iee.org.uk

End of exerpted webpage



-- Paula Gordon (pgordon@erols.com), May 02, 2001.


Interesting how the title includes the term "embedded systems," but the text contains possible problems in things like MS-Excel and Quicken. While most are technically embedded systems, this is not the definition of "embedded systems" which has been commonly referred to by Y2K-worriers.

-- Buddy (buddydc@go.com), May 02, 2001.

Excel and other data sorting programs get info spewed out of those little and not so little embedded systems.

If the data sorting programs say the bearing vibrations sensor's output is A-Okay, and its not then your power plant goes down after the scheduled maintenance was done. It takes a few weeks to get a bearing, the size of your car, made and delivered. If the information was known then an order could be placed, the part would arrive before failure, down time would be a few days at most - instead of weeks.

Was that a good enough example?

-- (perry@ofuzzy1.com), May 02, 2001.


You are missing the boat again Paula! There are two more crisis coming within the next 25 years or less that could spell doom to civilization as we know it....

1. We are running out of 9 digit social security numbers fast and 2. We are running out of telephone numbers in the current format.(xxx) xxx-xxxx.

Imagine the utter chaos that will reign in the streets of the modern world as systems crash and burn due to this!!!! Better save those preps a while longer kids.

-- Rob McCarthy (celtic64@mindspring.com), May 02, 2001.


Hey Rob,

Whats your problem? Seems that Paula must have done something to you in the past. Us kids seem to be doing alright. As I mentioned on another thread we don't have to prove anything to you. Actually you are starting to be a bit insulting.

Regards

Martin

-- Martin Thompson (mthom1927@aol.com), May 02, 2001.



Martin,

My issue is with the intellectual dishonesty she displays on theses boards. There are several important issues that we can be discussing, but no, we must discuss Paula's white papers on Y2k, a non event by every definition. If you think I was insulting due to my comments on social security numbers and phone numbers, sorry, but they are real issues too that will require another massive effort to fix. But, the world will not end either.

Regards,

Robert McCarthy

-- Rob (celtic64@mindspring.com), May 02, 2001.


Re-Post

Go Martin go. Tell em like it is. Only an idiot can't see whats happening here. All of a sudden we have the lights going out, refineries on fire, man holes exploding, pipe lines exploding, shortages in the pharmaceutical industry, retail stores that have been in business for 100 years are CLOSING DOWN. Basically the cost of ALL sources of energy have shot through the roof, should I keep going. I think not. Once an idiot, always an idiot. KEEP UP THE GOOD WORK MARTIN. YOU’RE THE MAN!!!!!!!!!!!!!!!!!!!!!!!

-- NdewTyme (NdewTyme@NdewT.com), May 02, 2001.


Go Martin go. Tell em like it is. Only an idiot can't see whats happening here. All of a sudden we have the lights going out, refineries on fire, man holes exploding, pipe lines exploding, shortages in the pharmaceutical industry, retail stores that have been in business for 100 years are CLOSING DOWN. Basically the cost of ALL sources of energy have shot through the roof, should I keep going. I think not. Once an idiot, always an idiot. KEEP UP THE GOOD WORK MARTIN. YOU’RE THE MAN!!!!!!!!!!!!!!!!!!!!!!!

-- NdewTyme (NdewTyme@NdewT.com), May 02, 2001.

Hey newdtyme, wake up and smell the coffee! You want us GI's to believe all these things have just happened in the last year?? Who's the idiot there?????? Life must be hell to live afraid of ones own shadow. Use your brain and join the rest of us GI's.

-- Rob McCarthy (celtic64@mindspring.com), May 03, 2001.


Dear Robert McCarthy,

Someone at the new TB2K site recently asked what purpose I thought my continuing involvement in the Y2K issue might be serving.

I have tried to be very clear about why I have a continuing interest. I have tried to spell out my concerns very clearly both on my website at http://www.gwu.edu/~y2k/keypeople/gordon and most recently in the postings at http://pub65.ezboard.com/ftimebomb200017873frm1.showMessage? topicID=245.topic . The specific posting on that thread that details these concerns is entitled: "Some thoughts (4/29) in response issues raised on...." 4/29/01, 10:45:17 pm.

In the event you did not read that thread, I will reprint one of my postings. I will only quote the headings that are discussed in greater detail in the 4/29 posting just noted reference above:

"I think it is useful to continue to discuss these issues for the following reasons:

1) Minimization of Future Impacts: It is possible to minimize the continuing impacts.....

2) Sound Public Policy: Policy needs to be based on understanding of the factors contributing to the problems....

3) Past Investment in Y2K Efforts is in Jeopardy: Failure to follow through on the efforts that were begun will jeopardize the investment of time, money, and human effort that was expended....

4) Loss of Understanding from Lessons Learned and Lessons Yet to be Learned...

5) Partial Efforts on Y2K Portend Poorly for Dealing with Other Threats and Challenges: Failure to complete the efforts that were begun portends poorly for the future success in recognizing and addressing future threats or challenges...."

Perhaps it would be helpful to add to this:

Since May 1998, I have had considerable first hand involvement with persons in key roles of public responsibility relating to Y2K, including key people involved in national and global policy. Some of them I had known before Y2K from FEMA and EPA and from other involvement in public policy issues in Washington over the past thirty or so years. I have also had a long and abiding interest in public administration, bureaupathology, groupthink, and organizational problems and other problems in leadership and management. I have a long time interest in how government can address complex problems, challenges, and threats and do so in a way that serves the public interest. To me public policies and government actions serve the public interest when they protect public health and safety and help maintain societal stability and a sustainable environment.

From the mid-sixties, I had been involved in many cutting-edge public policy issues. When I began to study the challenges relating to Y2K- IT and embedded system problems in May of 1968, I was struck by the way this public policy issue was being handled. Someday I may try to explain what I mean by that in a way that makes ready sense to those who may have no first hand experience of the public policy process and Washington. What I do have time to share with you now is that I thought from my initial involvement in this issue, that Y2K-related technology challenges were being handled poorly: the problem was poorly understood and defined, the options were poorly understood, the possible near term and long term impacts were not well understood, and a comprehensive approach was missing. Tremendous efforts were made nationally and globally to address the challenges. While they were extraordinarily successful in many ways, there were serious gaps. Remediation in the public and private sectors was by no means 100%. These gaps also have included failure to provide adequate monitoring, identification, and assessment of ongoing problems since the rollover. Based on the sleuthing I have done, there is no one involved in national policy or global policy ~ I repeat, no one ~ who is continuing to pay any visible attention to ongoing Y2K problems. If this is incorrect, I hope some one will come forward with names and contact information.

What the nation and the world went through can only be described as "Public Policy Russian Roulette". I have never seen anything like it in my years in and around government and government policy issues. We are very, very lucky indeed that there have been no Chernobyls and no Bhopals. I also think that we are lucky to be at a 4 on the impact scale and not higher.

Perhaps, I should also try another way to clarify the reasons why I think it is important that we have sound public policy.

No. 2 above reads: "Sound Public Policy: Policy needs to be based on understanding of the factors contributing to the problems...."

Let us say that you have pneumonia. You go to your doctor who misdiagnoses your symptoms and treats you for a common cold. Your condition worsens and complications develop and eventually a correct diagnosis is made and you are treated and regain your health, although it is a bit more compromised than it would have been had the correct diagnosis been made in the first place. Of course there are costs associated with all that you have been through as well. The question that arises in a public policy situation is this: What if the correct diagnosis is never made? and What if Y2K is (as was fully predicted and expected) a contributing factor in the problems that we are seeing across sectors in this country and globally? For those who contend that there are no problems, I can only say that if they were to study the GICC archives from the first half of the year 2000 alone, there is no way to escape the patterns of problems that have been occurring, all of the kind that were predicted and expected.

Some closing thoughts:

I think there is a better chance of addressing societal problems, challenges, and threats in an effective way if we are aware of all of their underlying causes and if we take all of their underlying causes into consideration in addressing them. To me that is just common sense. Those I have honored most in government over the past decades have exhibited such common sense. They have also had a commitment to serving the public interest. That is what I believe that people who serve in government should do.

It is not in my constitution to be quiet given what I have seen and given what I know concerning what went on and what is going on now. It is not my nature to be quiet when I think I can help clarify what has happened and what needs to be done.

Y2K and embedded system challenges constitute an unusual public policy issue in that it has been such a polarizing one. I am puzzled concerning how to approach individuals who seem disinclined to learn about what actually went on or disinclined to consider an alternative perspective. I continue to look for those in roles of public responsibility who do have an interest. I also continue to look for those technical experts who have broad based experience who are willing to speak on or off the record about what they know. I also am looking for individuals with expertise who are willing to speak directly to persons in roles of public responsibility, including individuals in the media, and help them understand the nature of the ongoing problems.

I hope that clarifies where I am coming from.

Sincerely,

Paula Gordon"

End of cross-posting from TB2K

-- Paula Gordon (pgordon@erols.com), May 03, 2001.


Miss Paula, I have read you far and wide, and I applauded your efforts when those in the "not acknowledged" sought to bring you down. I have also been in the same arena (actually sidelines) of those "supposedly computer experts". They did say the "Sky Might Fall", including Mr. Yourdan. I stumbled across a recent post of his where he poked fun at Mr. Gary North. I reminded him that he and his daughter had co-authored a book and the part section of the opening words were "Imagine if there were no electricity, and your ATM card won't work? I personally went through a year of hell and four ATM cards. Why should that happen after 15 years with the same card? I find it hard to stomach. Another of his sites (long past) said he only was trying to alert. Well "Alert" he did. So why now is he trying to shrink out and make Gary North appear the fool? Remember, you I.T. folks, we were depending on you for information/truth, and you failed, because you don't know what the Hell you are doing, you just draw a paycheck. There is no shame in not knowing, there Is Shame in not finding out.

-- MyStoryand (Iam@sticking.com), May 04, 2001.


To "My Story And..."

Thanks for your kind words.

I am not sure it is precisely the question you were raising, but I think that there are many reasons why so many have disengaged from this issue. For openers, the polarized nature of the issue can make discussing the topic very unpleasant. There can also be serious repercussions professionally as well as personally. A certain amount of stamina can be needed especially in light of all the disincentives and disinterest that currently surround the subject. While technical experts can talk with ease to other technical experts who share similar perspectives, it can be far more challenging for such experts to translate what they know for the benefit of a lay person. There is a need for technical experts who have this ability and there is also a need for technical experts who are able to communicate effectively with public officials and the press as well as with the general public.

We live in a time when those with technical expertise tend to be so specialized that they are not typically trained or accustomed to looking at the public policy implications of a problem. They may not be naturally inclined to focus their energies on trying to explain to public officials, the media, or the public what those implications are.

I think that we may well have reached a point where technology is threatening to "snooker" humankind. Technology has gotten incredibly complicated that few seem to even begin to comprehend it. The challenge is for those in positions of public responsibility to try to keep on top of this complexity and to do what can be done to ensure that technology is used to serve the highest purposes of humankind. This seems to be a central challenge of our time.

-- Paula Gordon (pgordon@erols.com), May 04, 2001.


Paula,

If you are still so concerned about this, then I would suggest that posting info. on such obscure internet boards as TB2K and GICC does nothing to air your concerns. If it is so important, then surely there are much better venues to use. Or is it that you've tried those and been ignored? That is the only explanation that makes sense to me. Therefore, I think your efforts have accomplished nothing since January 1, 2000. Sorry.

-- Buddy (buddydc@go.com), May 04, 2001.


Buddy, Paula is hoping to extract more leads so she can continue to grapple with and educate on the continuing "crisis". even though she continues to insist that people won't come out for fear of reprisals, she apparently has the confidence of people who will talk only to her on condition of anonymity. she is then able to press on, working behind the scenes, to whip this bug single-handedly. though how she is able to convince others in positions of responsibility to listen to her by citing anonymous sources, is beyond me. i guess it's a special talent?

-- bigwavedave (davidmoore01@exccite.com), May 06, 2001.

Exchange on energy problems on Downstreamer's discussion board may be of interest at

pub38.ezboard.com/fdownst...2672.topic

-- Paula Gordon (pgordon@erols.com), May 09, 2001.


Count me In. On how the supposedly "new and improved" ain't working. And someone, has to be accountable. Well, shall we count all the "new and improved" versions which take thrice as long for input? This was someones "Monster", and they should "own up". And a Hex upon their heads.

-- My Story and (Iam@sticking.com), May 15, 2001.


Moderation questions? read the FAQ