I found this clip from an article this afternoon

---Chevron won't fix all Y2K bugs in time

Chevron Corp., the fourth-largest U.S. oil company, expects to spend $200 million to $300 million to fix its so-called Year 2000 computer software bug, according to a company regulatory filing.

San Francisco-based Chevron said it has spent so far $40 million making sure its computers will recognize the year 2000. The company also said work won't be finished when the clock turns to Jan. 1, 2000.

-----------

Well, how the heck did they expect to get finished when they have only spent $40 million out of a projected $200 to $300 million dollars and there are less than 14 months to go! This sounds like gross negligence to me. The shareholders should be up in arms and a class action lawsuit should be under way. Less than 14 months and they have only spent between 13 and 20 percent of their projected cost.

-- Craig (craig@ccinet.ab.ca), November 09, 1998

Answers

Tava Corp has been hired by them for the embedded portion of the work. I have a "white paper" from Tava on the subject which says that they have found up to 20% of embedded devices in a few other companies are time sensitive (don't know the specific industry). This info from Tava was removed from their site the same week they signed the Chevron deal. Perhaps just a coincidence but everything is under scrutiny these days.

-- R. D..Herring (drherr@erols.com), November 09, 1998.

Craig - I agree, their announcement makes no sense - even if you try to add it up in Canadian quarters 8<)..

The first sentence says they expect to spend 200 - 300 million, then the next says they've only spent 40 million, but won't be finished by Jan 2000. So what are their options?

Figure some the budget is machinery and controllers - probably not a lot, but its a starting point. So the rest is salaries. They would need to spend 160 million - 260 miiion in salary in 12 months - given that Nov is half over, and people would not immediately begin in Dec.

At 50,000/paycheck per year, plus overhead and taxes doubling that cost to 100,000.00/person/year, it means they would need to hire over 2000-3000 programmers full time for all of next year.

They are so far behind they can't even spend their budget. And are they so far from finishing their audit so that they can't figure out whether they are going to spend 200 million or 300 million?

-- Robert A. Cook, P.E. (Kennesaw, GA) (cook.r@csaatl.com), November 09, 1998.

Hasn't been discussed on this board though that the biggest chunk of any Y2K budget is spent on testing?

Rick

-- Rick Tansun (ricktansun@hotmail.com), November 09, 1998.

I think in a perfect world you would be correct. However, I rather doubt that there will be much time spent on testing in any company that is so far behind.

-- Craig (craig@ccinet.ab.ca), November 10, 1998.

Robert:

2-3KiloProgrammers to be put to work for 1999 doesn't surprise me, consider howmuch distributed code there might be! In the BCBS plan I was a part of long ago, our piece of the over-all project planned out at 80,000 man-hours, in 9 months, which translated to about 60 people, and all we were doing is instaling a system that BCBC-USA (err-BCA) wanted each of its plans to have in house. This was about 40% of the project...........

I find it not unbelievable that the horn hairs ( as Cory would say) have not changed. "Throw more bodies at it and it'll be OK!"

C

-- chuck, etc (Driving@a.mess), November 10, 1998.

You're right, as usual, about the quantity of programmers - but I guess I was trying to imagine integrating together and de-bugging the individual routines separately being handed in from 2.5Kprogrammers all at different times, running from different versions, keeping up with de-bugging and re-programming results that affect the "future" users, etc. They would need to spend a considerable amount of time not programming, but testing their intergration scheme.

What? Me manage it? You want to try to control that bowl of spagetti logic?

-- Robert A. Cook, P.E. (Kennesaw, GA) (cook.r@csaatl.com), November 10, 1998.

Regardless of our "following the money" in hindsight, lets give credit where credit is due. Chevron is admitting, "we won't make it", which is a lot more honest than any of these other guys have been. It will be interesting to see if Chevron becomes the trendsetter, or if instead everyone continues to play this "we will make it" game.

BTW, this does provide a "living example" of verification that "following the money" is important. And how, because so little is getting spent, how little is getting done.

-- Jack (jsprat@eld.net), November 10, 1998.

I am glad Chevron is being forthcoming, but a few questions came into my mind. If these have been answered elsewhere, please let me know as I did not see them.

1. They will not be finished in time - Time being defined as 01/01/00 or some self imposed deadline?

2. If it is 01/01/00, what portion of the systems won't be completed? Are these systems mission critical or are we discussing 100% compliance?

3. Is this embedded systems or software?

4. Is it in refining? Transportation? Billing? Security? What?

I am not trying to be a Pollyanna about this, I agree it is a bad thing they will not be done. My thinking is though, just as we so thourglhy look at happy face reports for all the info we can get, we need to look at reports like this to get all the information we can get. What is good for the goose, is good for the gander.

Rick

-- Rick Tansun (ricktansun@hotmail.com), November 11, 1998.