All,

Do you envision the emergence, because of the ramifications of the fallout from Y2K, of a government sponsored approach to "Managed" I.S. This because of preceived threats to economy and national security a newly awakend government might forsee. For example, if left to themselves I.S. might implement another poor defacto standard, or miss a chance to impose one, that leaves us with a problem at some other future time. This view could embody state run or federal certification. Educational Standards and internships (it is still possible for people with less than a high school education to be employed as programmers and no standards are enforced). Or if there are other ramifications (perhaps a move by government at a high level to effect the industry through input into the language and compiler designs, thereby imposing government regulated and required specifications for date specific or other objects which may have infastructure or economic criticality. This would involve interaction between such companies as IBM and Microsoft with government bodies). Any input or discussion on this might make good subject matter not only for Ed's new book but also might help form a paradigm for the future so we are on topic. I'm not, for myself as a Sr. Analyst/Programmer, sure I would welcome such limitations at a low level (in other words I don't think the coders should be micro managed because of the stifling effect on individual creativity). However, at the abstraction layer of compilers, languages, and other things of this nature it might not be a bad idea. Nobody likes the idea of government involvement in anything because they can't do things in an efficient manner. They can however, in an albeight slow manner, ensure public safety. And, isn't that what this is all about? I mean hasn't this demonstrated that if left to it's own devices that this industry can effect the safety and well being of the people. Computers control the electric industry, the nuclear power generation, chemical plant and toxic waste and the transfer of money, goods and services, and international trade. If we have regulatory agencies for Banks, Electric, Nuclear, Environment, Agriculture etc.. Shouldn't the control mechanism which supports all those elements also be controlled at a very high level? I may be wrong, but it could take the art of Analysis/Programming to new levels of professional conduct. I realize the need for creative thinking, and I'm not talking about controlling or stopping that. I am only saying that through educational standards and certification, and better qualification of business rules built into the work benches of the developers we can eliminate problems in the future. i.e. Change management as a governmental/corporate collaboration in a kind of shadowing process if you that puts some intelligence in production.

-- Bill Lamoreux (wlamoreu@email.usps.gov), August 20, 1999

Answers

Bill,

This topic is exactly what I tried to cover in the draft version of Chapter 9 of my Humpty Dumpty Y2K manuscript.

My basic feeling is that if Y2K turns out to be a severe crisis, then society will decide that software is like atomic energy or various other inventions of mankind: incredibly useful, but potentially dangerous. Look at the way society has decided to license and regulate various other engineering fields over the past several hundred years...

Ed

-- Ed Yourdon (HumptyDumptyY2K@yourdon.com), August 21, 1999.

Bill: The worst possible thing that could happen to computer programming is for the gov't to have any control whatsoever over anyh aspect of it, other than in gov't installations. No one should have to have any kind of license or gov't sanction in order to be a programmer. Gov't involvement stifles creativity at the very least and controls operation and content eventually. With what we all know about the poisonous effects of gov't on nearly everything it touches, it's hard to build a case for gov't-imposed standards on any aspect of I.S.

-- cody varian (cody@y2ksurvive.com), August 20, 1999.

Cody,

I agree.. I wasn't stating that Gov't should manage the program shops. More that some involements with engineering standards to the venodors who own the tools for development. Like the development of the Visual C++ workbench (i.e. Microsoft). By controlling the tools you could control the really etremely important stuff like the Activex Objects in Graphical User devfelopment.. Another example would be those producers of Data Base engines.. Oracle has already done that with a Date (data type) but not because they were obligated to do so, only because of their own moral commitment. I think some things are too important to be left to chance. Bill

-- (wlamoreu@email.usps.gov), August 20, 1999.

I would respond that some things, in fact most things, are too important to be left to government. I see what you're saying but these things need to be worked out by the companies involved rather than by something imposed upon us by the government.

-- cody (cody@y2ksurvive.con), August 20, 1999.

Bill

By comparison, the US Government has used the FDA to regulate the development of new drugs to the extent that even in a time of booming biotechnological advances it is almost impossible to bring a drug to market...at least very expensive. This is all done in the name of "proving the drug first safe and then effective". Many good drugs never make it...many of the drugs that do still kill people.

I think that the US Federal Government will have dramatic drops in tax revenues for at least a few years and there may be a great era of "downsizing". Since getting the computers and embedded chips fixed will be one of the major hurdles to getting back to good tax revenue, let us hope that they will encourage a non-regulated solution.

It is my opinion that in a world which is evermore international, that government regulation will be taken over by private regulation similar to the ISO program. For example, there may appear in due time a certification company which is able to provide a service which audits software for any type of date rollover problems. Any software vendor who is worth dealing with will have sent the software over for review in order to get the "stamp of approval" before it is marketed.

-- Thom Gilligan (thomgill@eznet.net), August 20, 1999.

Bad as it is under our semi-free enterprise system, it would be only worse under more or complete government control and/or "licensing." EGTTTS (everything government touches turns to sh*t).

-- A (A@AisA.com), August 21, 1999.

Actually, it was a Governmentally "Managed I.S. approach" which precipitated the Y2K crisis!

It was the FedGov -- and the DOD in particular -- which established the two-digit date standard way back in the 1960s. The Washington Post ran an article about it about a month ago.

Obviously, this early exercise in "Managed I.S." left something to be desired.

"Those who cannot remember the past are condemned to repeat it." -- George Santayana, in 'Life of Reason

-- Stefan Stackhouse (stefans@mindspring.com), August 21, 1999.

Bill:

The software industry appears to be populated entirely by clever, greedy amateurs. I say this because if the industry had its act together, it would have set codes and standards for new software coming out in, say 1988, and we would not have much of a problem.

Instead, we have this sorry mess, which is not what one expects from an industry as profitable as the software industry. I can think of no other industry which could have gotten away with releasing and selling so many beta versions of code that were useless to customers. In other industries, it's called fraud. Used car salesmen, literally, would be prosecuted if they misrepresented their products the way the software industry does.

I expect considerable regulation of software by the government. If people will not police themselves, the government will do it for them. For those who are horrified by this prospect, just remember that it is government's enforcement of copyright laws that allows this industry to exist. Intellectual property rights have not been uniformly enforced over the years. The country is currently at a high water mark of enforcement, but that could change.

In short, I expect the very favorable legal climate that the software industry has enjoyed to become very chilly if Y2K causes a recession, and possibly some deaths. That will probably include much more than Bill has listed.

Dick Patton

-- Dick Patton (patton@ra.msstate.edu), August 22, 1999.

I would like to address Mr. Stackhouse on the issue that government precipitated Y2K through an early attempt at setting this standard. I have read the article in question, and yes their was an early defacto standard engineered in part by governement. This was not an enforced standard however. Since I have been a programmer from 1975, no governmental agency has enforced anything I have done. Programmers were always free to choose two or four digit years and were only limited by the capabilities of the toolsets provided by the vendors (and in some case not even then). Indeed I have coded year 2000 compliant code in all new applications I've written from that time. I agree that this could have set a general precedent, however government isn't to blame here, and indeed I don't think we are attempting to assign blame. We are looking at wether this may precipitate a recognition of I.S. as a potentially life/welfare threatening area by the general public and if Y2K will act as a catalyst for that. I believe it will. I believe the "reaction" will be yes, it is. I think we will have some post Y2K legislation in an attempt to set controls in place. Even as the banks are regulated by bank law to protect the investors. Again I am not nescessarily an advocate of low level regulation. I believe interaction at high levels with regards to steering the vendors providing the toolsets, and requiring an educational/intern minimum would be adequate to prevent a scenario like this in the future. However, I also do remember the old adage, "Once the camel gets it nose into the tent, the rest is not far behind", and another part of me thinks this may not be a good idea. So given this I though I would be an interesting post for discussion. In any event, I believe change is inevitable. I think even without any government interaction in this field, that management will be looking at this in a whole new way and the field will be forever altered because of this event. Please continue..

-- Bill Lamoreux (wlamoreu@email.usps.gov), August 23, 1999.

Even if Y2K is a "1", there will be regulatory efforts. From within various units of the present Administration, such efforts are already being discussed. While most of it is focussed on extending the capabilities for prying, there is also finger-pointing as to "how we got into this Y2K mess" and efforts to avoid similar future problems.

The apparent thinking seems to be a mix of law enforcement needs and wants, fear of encryption, and pure and simple governmental control of programming and programmers...

Just business as usual...

:-), Desertrat

-- Desertrat (arthur@surfsouth.com), August 23, 1999.

To Bill Lamoreux:

My point was merely: just because you have a government standards setting mechanism, that is no guarantee that the RIGHT standards will be set. I realize that the 2-digit standard was only enforceable through government contract specifications, and thus was only a de facto rather than de jure industry standard. The fact remains that in this early exercise in government IS standard setting, the performance was not exactly "confidence building", shall we say.

We have the ANSI structure already, setting a vast array of standards for a wide number of industries and disciplines. For the most part, these are well established and seem to work pretty well. In many cases, government involvement is limited to incorporating various applicable ANSI standards into its contracts, and that is enough to establish the standards as de facto. Insurers through UL are another major force in making ANSI standards de facto. I would guess that the main reason that ANSI standards are so universally adopted, however, is simply because the engineering and other professional or trade associations are intimately invloved in developing the standards, and so there is a high degree of collegial confidence and buy-in.

Rather than "re-inventing the wheel", what we really need to be asking ourselves is: "Why is this "wheel" -- which works so well for other industries and disciplines -- not yet working well for software?"

-- Stefan Stackhouse (stefans@mindspring.com), August 23, 1999.

Mr. Stackhouse,

I agree that the wheel doesn't operate correctly. It hasn't ever operated correctly. Basically this is an old argument going back a long time. Governnment had the opportunity in the past to set a good standard, and instead they set a precedent which alluded to the correctness of time and a 2 digit year. They dropped the ball there end of story. However, industry as a whole has also dropped the ball. It didn't react to this situation well or in a timely fashion. It would be nice if a consortium of industry would collaborate on the really important and fundamental issues. There have been some successes, case in point, ANSI as you pointed out. However the industry will be judged not by it's successes in this case, but rather by it's failures. Before the Challenger blew up there had been what, say 20, successful missions. Never the less, such a catastrophic failure was so extreme that the public opinion would have turned against the NASA agency, had not a thorough review, determination and new standards there would have been intervention by a higher governmental agency to abort and further control the agency. Because of the nature and financially strategic advantages offered by new software, it is the nature of the industry to release at the earliest opportunity and create standards to undercut competetion. Although this has led to tremendous advancement, a catastrophic failure will indicate the need for some change. I predict that if there is no catastrophic (life threatening) failure directly attributed to Y2K that there will be no change in the industry. Simple business failures will not be enough to convince either myself or the public that we need regulation. If however, we see some massive failure in the supply chain (food, medicine etc.) or chemical plant failures or other such "China Syndrom" attention garnering failures, it is inevitable that some action will, and should, be taken. Given that, I reserve judgement and will allow myself to be influenced by the outcome of the Y2K event. Never before has any nation made such a concerted and synchronized effort, under mandated management controls such as the IV&V process. If we can't get it done right, given the time, resource and change management effort employed then perhaps we do need to slow down. Bill

-- Bill Lamoreux (wlamoreu@email.usps.gov), August 26, 1999.

Ed said ...Look at the way society has decided to license and regulate various other engineering fields over the past several hundred years...

Ed, I think that it is more likely that some official somewhere decided that the scope of control of his organization would increase if certain activities were regulated/licensed and sold the idea to the public. Most organizations want to extend their control over their environment/destiny -- that seems to be their reason for existence in many cases.

George

-- George Valentine (georgevalentine@usa.net), August 26, 1999.

From: Y2K, ` la Carte by Dancr near Monterey, California

There is more discussion of this question at Ch. 9 - Commentary.

-- Dancr (addy.available@my.webpage), September 07, 1999.