Early on, y2k struck me as pissing down the rathole. I was under the impression that vast sums were being (or would be) spent to do no more than tread water, making software do nothing more than it had ever done. Further, I was convinced that this effort would fall short, leading to a world of downtime, shortages, delays, and all manner of inconveniences. And that these would add up to even greater cost. Finally, there would be a huge wave of lawsuits, causing more huge sums to switch hands from one company to another, except for the large percentage that ended up in the lawyers' pockets. A total lose-lose-lose proposition. My ex-rectum breakdown of costs was 70% lost productivity, 20% legal, and 10% code fixing.

In hindsight, this was a really LOUSY analysis, perhaps growing from my concern for my own personal safety. I should have realized that it's an ill wind that blows nobody good, and there's rarely such a thing as a pure cost. At this point, I'm swinging the other way and wondering if we have ended up with more silver lining than cloud.

Of course, a large part of this swing derives from the observation that lost productivity has been trivial and therefore legal action has been minimal. Even the most ardent optimists (NOT counting the "evil polly" contingent that did nothing but hurl insults) were astounded that we have had NO significant documented y2k impacts, anywhere in the world. I admit this still confuses me -- I expected y2k to cause at least *something* to go very wrong somewhere.

Be that as it may, I think we need to examine the code-fixing expense (that being all there was, except for the net effect of the hundreds of small and isolated minor glitches we've had) to see what we spent and what we bought. Was it a bargain? Consider:

1) Modernization. Upgrades have been rampant, both of aging and obsolete hardware and of long- overdue software upgrades. Many organizations were down-rev a large number of revisions. Coming up to the latest revision sometimes (how often?) required new hardware to run the new code. The result is more robust code running faster.

2) Cleanup. In the remediation process, programmers found all manner of non-y2k bugs, especially in error handling routines rarely if ever exercised. Generally these were fixed (or at least flagged) in the process. Programmers also found great gobs of dead code, unreachable by any logic path, that had been left in by prior generations of programmers. I've heard of cases where they found whole systems nobody had used for years -- some of which wouldn't even RUN under the current OS. And countless obsolete reports were being generated that nobody ever read.

3) New implementations. An estimated 25% of large organizations elected to throw out legacy systems where barnacles had been growing on other barnacles for decades, and institute whole new ERP systems like SAP or Peoplesoft. This is an expensive proposition for many reasons (software cost, implementation cost, changes in the way of doing business) but SAP etc. have been successful because once the pain has subsided, the patient thrives.

4) New architectures. Some organizations switched from mainframe-centric to client-server, from centralized processing to department-specific, from "handed down from God" decision making to localized decisions, which tend to involve less paperwork and be more flexible, more appropriate, and a lot quicker.

5) Documentation. Shouldn't be underestimated. Many organizations really didn't have a handle on what code they had, what it all did, or even who was responsible for what. In well-managed shops, y2k was the impetus for documenting code, interfaces, systems, inventories, licenses, versions and all whatnot. I know from my own experience how very valuable this can be. Some outfits didn't even have source code to some of their systems, and many had no idea which (of many) versions of the source was actually running. This has changed.

6) Training effect. In some shops, both the depth and breadth of understanding of the code base has been increased, for both programmers and managers. Hopefully, this will enable programmers to work more quickly and introduce fewer problems, and enable managers to make more informed decisions.

Of course, NONE of these is a pure benefit -- all were side-effects, spinoffs from the primary purpose of the remediation exercise. And all had costs, monetary and otherwise. And many of these benefits are long-term, not yet fully realized. We'll probably never know how much of the total remediation expense would, in the final analysis, have been worth spending even if y2k had never happened. But I think there's no question the money wasn't a total waste, and viewed broadly, we're better off for having spent it. Was it a bargain at the price? I don't know. Do you?

-- Flint (flintc@mindspring.com), March 09, 2000

Answers

Funding was allocated where it never would have seen the light of day. For that, we will learn to be forever grateful.

-- Sifting (through@the.rubble), March 09, 2000.

Flint

Having spent over 2 years on this Y2K thing I'd have to say it was very beneficial. Migraines and nervous breakdowns galore preparing for audits but well worth it.

A lot of new equipment and cleaner code. Hopefully we won't wait so long to upgrade and maintain the next time around.

Was it a bargain?? It's difficult to term 'billions of dollars' as a bargain but it is probably better than the alternative - nothing is working.

Deano

-- Deano (deano@luvthebeach.com), March 09, 2000.

I think it was beneficial, and definitely NOT money wasted. The time was long past to look into systems that had been running for years with no review. At one site, I was simply asked to look into the remediation process required. I think I found 3 programs out of 150 that needed changing and probably 70 that hadn't been used in 5 years or more. The client asked that I delete the old code, which was just taking space in the library system.

In addition, I feel it provided the impetus for many firms to move forward in technology where previously they'd felt "comfortable" in simply having something that "worked."

You didn't mention the impact on non-technical folks who were influenced by Y2k, but I will. I think several folks on the old TB2000 forum were positively influenced by Y2k. There were some who had always wanted to live in a rural environment, but just couldn't find the excuse to implement that dream. There were others who found themselves living "day to day", and learned something about planning ahead and the savings afforded by so doing. Of COURSE there are OTHERS who resent being "taken", but life doesn't hold much time for resentment. They've learned and have moved on.

No side of a coin exists without the other, however, so where SOME firms moved forward and beyond, the firms who chose NOT could find themselves out of the market forever. Legacy programmers have suddenly found themselves out of the job market. Some non-technical folks feel they OVER-reacted and don't like the choices they made. As with the legacy programmers, they can either move on, or blame others. The choice is theirs.

-- Anita (notgiving@anymore.thingee), March 09, 2000.

Well, at my site we did an analysis (of the SMF log tapes for any MVS types reading this) in mid '98 to determine all programs which had run in production the past 18 months. As a result we moved over 3,500 items (Progs, JCL etc) which appeared not to have been used to libraries designated as obsolete. So far we've had to pull about 10 (mostly ad-hoc SAS jobs) back to prod libraries again.

This clean up was about 20 years overdue and would *never* have happenned without Y2K.

RonD

-- Ron Davis (rdavis@ozemail.com.au), March 09, 2000.

Once again. Too much to read unless I was in love.

-- Carlos (riffraff@cybertime.net), March 09, 2000.

Was it a bargain at the price? I don't know. Do you?

No real idea Flint, but I suspect from a business standpoint there are winners and losers and no accurate summation will ever be available.

Personally, the mental prep and purchasing of insurance was a valuable exercise in that I collected a bit of wisdom which is always usefel, regardless.

-- Will (righthere@home.now), March 10, 2000.

Not to forget benefits in the area of testing. Vast testbeds/harnesses had to be set up to provide system testing facilities. Apart form Y2k it was unusual ~(not unlnown) to have to test entire systems when implementing changes (as y2k was all pervasive).

These test facilities (files/jcl/software tools/database extracts etc) are now availble for new projects.

-- Sir Richard (richard.dale@unum.co.uk), March 10, 2000.