http://www.msnbc.com/news/289588.asp

-- (abc@def.gov), July 14, 1999

Answers

I hope so. I really do hope they make it. Their task though is huge.

The agency has been sharply criticized over the pace of its computer repairs. A report last month from the Office of Management and Budget said 264 of the departments 2,096 most important computer systems still were not compliant  more left to repair than any other federal agency.

And Congress General Accounting Office said last week that Defense experts face significant problems and that the agency was far from successfully finishing its various Year 2000 end-to-end test activities.

-- (hopeful@but.watching), July 14, 1999.

I notice in the article that this was a test of only 44 "systems". And yet Koskinen says:

``What we're seeing is the ultimate in testing,'' said John Koskinen, chairman of President Clinton's Year 2000 commission.

Here's another version of the article: Washington Post article

And yet another version of the story that - as Paul Milne points out on c.s.y2k - shows that this test was of 44 out of 1000 systems:

DOD puts vast nets to Y2K test

So... they tested 44 of their MOST CRITICAL systems - systems we assume they had spent the most time remediating and separately testing. And still they had glitches in 3 systems. What about the 936 other systems?

And what does this mean that this is the "ultimate in testing"? This was it.. finished testing. Wrap it up boys. We've got our story. ???

936 other systems not even addressed by this test.

And what do you make of this statement:

Technical experts built a duplicate network -- what they called a ``parallel processing environment'' -- then rolled those computer clocks forward to simulate the week following Feb. 28, 2000.

and

The computer network being tested is a duplicate of the Pentagon's genuine supply system except for the volume of requests running across it, Goldstein said.

Doesn't sound like quite a full - or ULTIMATE - test at all, does it?

Any thoughts?

-- Linda (lwmb@psln.com), July 14, 1999.

Right you are Linda. They are limited by their management's ability to manage, their staffing resources, time and the complexity and antiquity of their systems. One could see this as a vast 'pruning' of IS/IT systems (alot of wood will be hitting the floor never to be recovered) and possibly a single stem with a few leaves remaining. They only tested a few pieces of that single stem in a mock environment (not the real thing).

They are doing the best that they can within their limits. That said, we must ask if any large organization will be functional afterward. My guess is 'no'. This will rot out the orgs and many will fall apart slowly as they can not deliver goods or services one way or another (for whatever reason).

Its not that these orgs can not operate with a paper system. They could if they knew how to and had the staff to do it (and supplies). Its that all that stuff was replaced with computers. Subsequent increases in volumn and effeciency were built into these orgs using computer technology.

The computers are the bones of the system and the information is the blood. The telecom and data interfaces are the circulatory system. We are headed for a major trauma to these systems, like a body impacting a hard object. What will be broken and will the 'corp' survive? Trauma leads to shock and shock leads to death in some cases.

That they are testing so few in a testing environment as well as calling it 'the ultimate test' means to me that that is THE level we can expect at the most in that kind of situation. They will Fix On Failure when they throw the switch on the renovated systems, which means FAILURE and then alot of agnoizing over how to approach the scene of devastation.

-- ..- (dit@dot.dash), July 14, 1999.

Trying to keep an open mind, but see several problems:

For instance as above "A DUPLICATE SYSTEM"? I have hard time believing most people wouldn't create a "duplicate system" designed to WORK, no matter what the original was like. Not to mention 'how do you create' a duplicate parrallel system, that took years and years to develope, in a few weeks or months?

Of course, it was just a mistake we bombed the Chinese embassy, and ...oh yes ........ estimates of atrocities and success in boming missions in Yugoslavia were of by a FACTOR of 10 (1000% off?)....a minor glitch, or was that 90,000 minor glitches............

-- Jon Johnson (narnia4@usa.net), July 14, 1999.

Ok I'll be nice (I was going to throw flames but I understand that you really know nothing about IT projects). This is the way testing is done. Don't look to closely, you might go blind. Testing can not be done in the production envrionment. It almost lead to WWIII back in the seventies and it hasn't been done that way since. Before an application goes into production you must first develop it in a development environment then test it in a test environment. These two can be the same but that tends to slow development, so if you have the bucks you build two separate environments. Hopefully, it's a duplicate but again depending on the bucks, you may just have a mini version of your production environment.

Think about it guys (after all you say you can "see" the interconnectedness - OK that was a little sarcastic). How could you roll a clock forward on your production environment? That would mean total disruption for all your interfaces operating in today's date.

-- Maria (anon@ymous.com), July 14, 1999.

Maria, that happens to be the point. NOT that you don't need to test the way you stated, but should the DOD crow that everything is beautiful? Because there is a margin of error, I can believe the systems they actually tested in this way LOOK positive, but the "Y2K is OK" spin is way premature, IMHO.

-- Jon Johnson (narnia4@usa.net), July 14, 1999.

Maria is correct. Really smart testers are good at knowing what needs to be simulated and what can be left for production fixes (the infamous FOF, which is a legitimate process, when properly bounded). There is always some "crossing of fingers involved", but the trick is to keep it to manageable levels.

I would add parenthetically that one of the biggest reasons I am a Y2K "doomer" is because I believe we are effectively trying to simulate "testing" of the entire world systems/infrastructure, an inherent logical impossibility. And, obviously, re-introduce all those systems within a tiny window of time (even six months a tiny window).

It "may" work, but prep=hedging bets for my family.

-- BigDog (BigDog@duffer.com), July 14, 1999.

Now that DOD has tested 4% of their 1000 mission critical systems, how long do you suppose it will take them to complete and test parallel environments for the other 96%? Five months???

-- RUOK (RUOK@yesiam.com), July 14, 1999.

"Zach Goldstein, director of logistics systems modernization in the Office of the Secretary of Defense, said the logistics systems test, under way at the DOD Logistics Y2K Operations Center in Fairfax, Va., focuses on just 44 of the 1,000 DOD logistics systems."

This test was only of their LOGISTICS systems, and 96% of their LOGISTICS systems were NOT TESTED. Got that Maria?

-- a (a@a.a), July 14, 1999.

Thanks a, but I can read and understood that point perfectly. But what's your point?

-- Maria (anon@ymous.com), July 14, 1999.

Hmmmmm, let's go back in time to the media coverage of the Gulf War in 1991.

"Our Patriot missles have hit almost every SCUD launched, and our smart bombs have not missed a target yet."

Which became:

"Our Patriot missles missed a couple scuds, and our smart bobmbs have been ninety-something percent accurate"

Which evolved into (some years later):

"We couldn't hit the broad side of a bull's ass with the Patriots or the smart bombs. Sorry about the barracks. Sorry about those civilians, etc. etc."

Obviously this is an exaggeration, but indicative of the DOD's ability to manipulate the media. Of course they wouldn't do that now, would they? :)

-- ariZONEa (pinocchio@dod.gov), July 14, 1999.

Got that Maria?

-- a (a@a.a), July 14, 1999.

The sad part about the reporting of this test is that: a limited scope test in a duplicated environment that covered only 4% of their logistics systems to a limited number of fleet and Army units is being help up as an example of the "most rigorous system testing ever" that proves "our war fighting systems will work next year!"

- and you know what? It IS the most rigorous integrated system testing ever done by the government! That is the sad part.

-- Robert A Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), July 14, 1999.

Hi folks... I'm back. Let's get down to brass tacks: DOD and John Koskinen are taking what does indeed look like a qualified success and is certainly good news, and turning it into a proclamation that all will be well. They are extrapolating big time from the success of a minor test and asking you to do the same.

Maria, you seem to make the doomer argument by saying it's impossible to test fully, in real time. This is what Yourdon and (I shudder to utter the names) North and Milne have been saying for years. You draw the complete opposite conclusion from the same assertion. Wonder how that happened?

Cheers to all!
Scott Johnson
Editor,
y2ktoday

-- Scott Johnson (scojo@yahoo.com), July 14, 1999.

Scott, sorry to burst your bubble but this is the way it's been done for years. Do you know what DT&E and IOT&E are? Nothing new here, move along. But then I'm supposed to jump to the conclusion of a 10 scenario.

Ok Robert how many government test have you been involved in to conclude this is the most rigorous?

-- Maria (anon@ymous.com), July 14, 1999.

Maria -- Yes and no. As I said, I agree with you that all testing is done in way we've been describing (no secret to Robert and a, to say the least). Actually, I was in a rush and didn't realize how minor the tests were until rereading this thread. You don't seriously quarrel with that characterization, do you? Rough rule of thumb: they should be testing AT LEAST a third to a half of all systems.

-- BigDog (BigDog@duffer.com), July 14, 1999.

The article said, "It involved 44 military computer systems and more than 1,000 people in 22 locations, including five Navy ships." I'm not sure where the 1000 systems come from. Maybe a can explain.

In any event, are trying to say that this proves a conspiracy or a 10 scenario? It's an article talking about the success (only 3 discrepancies) of a test. What are you trying to conclude? I conclude good news but I was never really concerned about the logistics. IMO, logistics nevers comes close to critical in the DoD world. The modernization of these systems will take a long time (a turtle's speed) and they are just beginning with the just-in-time inventory. NORAD has a thirty day supply on hand at all times, so those systems don't depend on logistics. That doesn't sound too much like jit.

-- Maria (anon@ymous.com), July 14, 1999.

Maria... nowhere did I say you should assume a 10 scenario... I don't buy that myself. But to compare "how things have been done for years" to the largest single technology project in history, geared around preventing roughly simultaneous failures around the world, is folly IMHO. Tests have failed for years, but never has this kind of situation been tested for.

By the way, I've lurked here for about a year and posted occasionally, and I've read many of your posts with great interest. You're obviously intelligent, articulate, and know your stuff. But you also make the same mistake that many "doomers" (and "pollys") do, which is to take any piece of evidence, no matter how inconsequential and/or inconclusive, and use it as incontrovertible evidence of your position. Bad idea, IMHO. Nobody knows what is going to happen, and a lot of people are going to be hugely mistaken one way or another. I think you should at least allow for the possibhility that you might be in that category.

Cordially and with respect,
Scott

-- Scott Johnson (scojo@yahoo.com), July 14, 1999.

Mr. Cook wrote, " and you know what? It IS the most rigorous integrated system testing ever done by the government! That is the sad part."

It's really sad because it shows the incredible size of the entire military Y2k project.

Maria, think about it.

The very size and scope of the problem for the DOD is equal to a monolith. Only 44 out of 1000 logistical systems tested = the largest testing ever done by the military. That is amazing.

Can the goal of testing all systems be met realistically?

Can all the remediated systems be put back into service in the next 5 months, realistically?

I know that in a perfect world it might be possible.

But when did a perfect world ever exist? Realistically, I mean.

They might have done developing new technology and software instead of trying to fix what they had built upon for almost half a century.

This is good news?

I read about this story last night at about 11:00 pm pst.

This test started in the morning of 7/13 and ended on 7/13

How can a test - as extensive as this was supposed to have been - be completed in hours, realistically?

I'm clueless but I'm firmly grounded in reality. I want good news but I want the news to have substance. Stories like this only deepen my mistrust of what the government says is good news.

This is government spin based on fantasy, IMHO. They might as well have started off the story, "Dear Penthouse, this kind of thing never happens..."

Mike =====================================================

-- Michael Taylor (mtdesign3@aol.com), July 14, 1999.

Sorry Scott but I didn't comment on the article (it didn't mean anything to me). My first post was about testing in general because I thought that Linda needed some explanation. I didn't conclude things were great or that things were bad. I agree that this situation has never been tested. And I expect problems but not as much as most on this forum.

My last post represents my opinion of logistics command in the DoD. Again the article states they've tested; just mark it up as another data point to draw some conclusion in conjunction with other data points.

My comments on the 10 scenario were specifically for a who does believe in a 10. It seems he wanted me to comment on something but I still don't know what.

Thanks for your positive comments about me. Most here don't feel that way.

-- Maria (anon@ymous.com), July 14, 1999.

"Thanks for your positive comments about me. Most here don't feel that way."

You're quite welcome. There are many eminently qualified people on this forum (and in the "real world" as well) who disagree completely on the likely scale of the Y2K event. After almost a year of working on covering the issue, I honestly don't feel like I know much more about what's going to happen than I ever did, but a great deal of what I have learned, I've learned by watching the debates here. I have also learned to sift out the crap on both sides of the debate and focus on the substance. It's quite a useful skill, IMHO. :)

Scott

-- Scott Johnson (scojo@yahoo.com), July 14, 1999.

Neither a nor myself believe in a "10". Expecting a depression is not a 10. As I recall, a rated the chances of a "10" at 1 percent or so. 100 to 1 is not a 10"!

-- BigDog (BigDog@duffer.com), July 14, 1999.

Although the analysis there isn't as good as the analysis here (at least on some threads!) Y2KNEWSWIRE wasn't blown away by the DOD testing, either. Crunch this

-- Dog Gone (layinglow@rollover.now), July 14, 1999.

Maria,

The TRUTH is... HERE

-- cb (.@_._), July 14, 1999.

Dog Gone (it):),

You beat me to the link. As always, Good to see you here and the 'republic.

-- cb (.@_._), July 14, 1999.

Sorry to correct you BigDog but a said on another thread (unless it was an imposter) that he believed in a 10. Well ok then substitute 10 year depression for a 10 scenario.

Mike, my response to you: the size of the military project is great no doubt about it. But Y2K is not a difficult problem technically; it's a tremendous management problem. Now believe it or not the military invented project management and most corporations are starting to adapt these processes. So the military does have some good background on how to run a project. How do you eat an elephant? One bite at a time. Same with PM. Each commander is responsible for that organizations applications and reports up the chain. It can be done I tell ye. (Sorry but I haven't found the 1,000 systems yet only a's post in italics and that doesn't count, so I won't comment on what has been done and what hasn't been done. Point is I don't know (and don't care) and you don't know.)

Yes the goal of testing can be met realistically. Again you take it one step at a time in a commander's organization. Back to the technical problem. Y2K doesn't change functionality (my opinion, it's more like maintenance), so you check for the changes you made, specifically the dates and some regression testing to ensure you haven't screwed something else up.

Systems are not going into production in the big bang method. As they become remediated, they go back into production. Then you can take a "snapshot" of the production environment after all is remediated and run an integration test. Yes that can be done in five months, it can be done in two months depending on your environments.

Please explain "They might have done developing new technology and software instead of trying to fix what they had built upon for almost half a century". Yeah so production is down.

Extensive depends on your definition. According the these folks it means involving 44 system, 1000 people in 22 locations. That sounds pretty extensive to me even if it did only take one day. (The article lead me to believe it lasted a week but can't recall). Now rigorous (in my mind) would include number of test cases and fault (negative) tests. But going back to Y2K, functionality doesn't change and rigorous is in the eye of the beholder.

Doomers say you want good news but yet you continue to pick apart each and every word of any good news article. Just take it for what it's worth and don't let me stop you from preparing. I won't be able to change your mind and I dont' want to be accused of killing people because I told them not to prepare. Sorry for the long post.

-- Maria (anon@ymous.com), July 14, 1999.

Maria, yes I believe in a 10, if you mean 10 year depression. And I know good and damn well you hang on every word I post, so you know exactly what BD is talking about.

DOD's task is overwhelming, I wish them luck, and that's all I can say without jeopardizing national security or risk being accused of sedition by Decker.

And do your homework, honey. The 44/1000 was a direct quote from the director of logistics systems modernization in the Office of the Secretary of Defense, in the article at

http://www.fcw.com/pubs/fcw/1999/0712/fcw-newsDOD-07-12-99.html

If you read Paul Milne's posts on c.s.y2k, you'd have known that. :)

-- a (a@a.a), July 14, 1999.

Here maria, more "good" news:

Technical experts built a duplicate network - what they called a ``parallel processing environment'' - then rolled those computer clocks forward to simulate the week following Feb. 28, 2000. ``What we're seeing is the ultimate in testing,'' said John Koskinen, chairman of President Clinton's Year 2000 commission.

Yeah, right.

-- a (a@a.a), July 14, 1999.

Maria, first, thanks for your response : )

>>the size of the military project is great no doubt about it. But Y2K is not a difficult problem technically; it's a tremendous management problem.<<

We agree. And management doesn't have a great track record so far. That's exactly why I made the comment about the "perfect world" not being realistic.

When I should have written was, "They might have done [better] developing new technology and software instead of trying to fix what they had built upon for almost half a century"

What I meant is that "necessity is the mother of invention".

That is to say that perhaps if the problem had been looked at early on and assessed as being more than just a "management problem" maybe new systems and processes could have been developed.

It's my understanding that many of these systems run virtually on top of each other in code layers written long ago in languages no longer utilized. Some of these systems go back decades. Well, I would think that technology and code writing have advanced since that time. Certainly, hardware has improved.

If someone was developing a new OS for a computer which runs on the most cutting edge of hardware technology they would write the code to run optimized for the hardware in that system. If someone is writing an application to run on that hardware with that OS they would create the application in code that would take advantage of the processing power of the system.

Forgive me if I make errors in the following analogy and be kind : )

Y2k reminds me of what Intel is doing with the Pentium and the Merced chips. These chips are required to be larger, run hotter, use more power than the Motorola/IBM G3 and G4 chips because Intel has to include outdated instructions and functions in order to work properly in a Wintel box. Why? Because these outdated instruction sets and functions are required in order for a Wintel computer to run properly. The old code is still in there, required for the OS to run on the system.

What am I trying to say? What *AM* I trying to say? : )

Well, the DoD is fixing systems that perhaps should have been dumped years ago because better technology now exists. They aren't pushing the envelope - they are only spending money to continue operations and avoid failure. Billions of dollars could have been spent in stimulating and further advancing our technology, perhaps to the point where the human culture advances and a DoD is no longer necessary.

Instead, we are asked to be happy because 44 out of 1000 systems are claimed to be peachy in an entity which, by shear scope and size, boggles the mind.

Hurray. Personally, I think we lose either way.

However, in the end, you are very right. I don't know.

Mike =====================================

-- Michael Taylor (mtdesign3@aol.com), July 15, 1999.

See also:

Link

-- CD (not@here.com), July 15, 1999.

Ah Mike you hit on something that we all wonder. Having worked for the gov in one form or another for some twenty years, I think that bandaids was pretty much the mode of operations for a long time. However, (before Clinton got into office) that changed and the gov started to overhaul its systems. I'm not that close to logistics systems but they started a "modernization" back in 1992. Now with Clinton, the funding was cut dramatically to mostly just putting in bandaids once again. Fortunately, the DoD had a good baseline of improved systems, but in eight years technology moves quickly and now they are "slow" and "out-dated".

If you're talking about business, I think that corporations (mine as an example) are trashing the old stuff and not just fixing the code.

Just a quick comment on management. Yes it doesn't have a good track record but it's improving tremendously. Go out to some web sites, you'll find that PMs get $100/hour and the DoD invented the profession. Really the DoD follows many processes that corporations are finding valuable and are adopting. That's my background. I think you're just repeating the Yourdon chant. True it may not be moving that quickly but in order to be competitive, corps need to stay ahead of the game. These processes help give that edge.

-- Maria (anon@ymous.com), July 15, 1999.

no, 'a', it's I who hangs on your every word.

BTW, Maria, good to see you back and obviously well-rested.

-- lisa (lisa@work.again), July 15, 1999.

Maria,

You're right again. My opinion is based on working in corporate situations and not for any government entities and I use a computer out of necessity not because it I ever intended to. Heck, when I started out as a graphic designer the Macintosh didn't exist.

I can only take what you write and hope that it's true. One of the reasons, perhaps the second most important reason, that I became self-employed was because of the many times I bumped heads with clueless VP Managers who cared more that I fight freeways to get to work by 9:00 am than that I stayed at work until 9:30 pm the night before to finish a project. IMHO, it isn't wise to call an employee into an office after they've just driven 25 miles to work in heavy traffic, and rail against them for being 5, 10, 15 minutes late. And then, to have the VP Manager threaten to dock your pay (even though you are salaried) because you are late.

"(thinking...uh, duh) Salary means that you work on a day rate Mr. Manager which means you're paid if you work 15 minutes or 15 hours."

Mr. Manager didn't like me much after he went to HR to try this and realized what I had told him was absolutely true. Why did I know it was true? Because the same situation had happened at every single work situation I had been in prior. Good quick tip, always become friends with your HR department.

So, I'm a little jaded about *some* management out there. My perspective IS skewed toward the negative. Some managers were great, others totally clueless newbie college grads in their very first position who become VP of Marketing because of nepotism. I have so many horror stories!

So, what the heck is my point?

Working in the real world we get to come across the best and the brightest but they're mixed in with the worst and the totally clueless. Some will inspire their employees to do great work, others will make more problems for a company rather than solve them.

And, Y2k remediation is happening in the same world that created the problem in the first place.

When I read that 44 out of 1000 logistic systems are tested (which actually represents only minute, almost nill, portion of the entire project) and Koskinen crys "success!" I see a manager who is very, very interested in telling his boss, "Hey, we're doing great! No problem!"

I hope he's right but based upon what's been completed so far can he really make that assessment, realistically?

Maria, I really appreciate your perspective and the insight you have. Thank you so much for taking the time to respond! I always look forward to reading your posts and I really do hope you are right!

Mike ===================================

-- Michael Taylor (mtdesign3@aol.com), July 15, 1999.

Mike, you hit the nail on the head there.

BD, Neither a nor myself believe in a "10". Expecting a depression is not a 10. As I recall, a rated the chances of a "10" at 1 percent or so. 100 to 1 is not a 10"! It seems to me that a lot of Pollyannas lump everybody who disputes the BITR scenario into the collapse-of-civilization crowd. It's easier for them that way: they don't have to think, they can just react to an extreme position.

-- Lane Core Jr. (elcore@sgi.net), July 15, 1999.

Maria, Scott & Micheal Taylor,

I've enjoyed your discussion, but while you have been giving serious thought to the Testing segment of our DoD's logistics systems, someone in Management just told a crew thats been working on a NEW application for The DOD that would either improve their systems of assist in testing, etc. Was CANCELLED after they had just completed, because there wasn't any MONEY for the Implementation of the Funded Application.

Sorry.

Start on a New Project, boys. You just wasted gobs of Money and a 6 Month period of usefull time. But, we've lost our Funding now that we are ready to launch the product. Oops.

Now, that's Practical Government Management.

Kind of makes the point of Testing Moot, doesn't it?

And it happens more often than you think. Not to mention that Management still doesn't get it.

Think of Senator Grahm, from Texas, (sorry bout the spelling) when you think of clueless Management.

And lastly, Maria. You are indeed inteligent, but you are also likely vastly more intuitive, understanding and competent than any Manager you've worked for.

The thing is your management still doesnt' have a clue, the money, the time, or the inclination to do what is required.

Father

-- Thomas G. Hale (hale.tg@att.net), July 31, 1999.