Mindre än 400 dagar kvar... har Ditt företag inte gjort något ännu åt år 2000-problemet är det lika bra att börja leta efter ett nytt jobb. Men det finns fortfarande en LITEN möjlighet att klara det om företaget ännu inte startat: Sätt ALLA tillgängliga resurser på problemet - kosta vad det kosta vill. Visst, det kommer att kosta många kunder, men företaget har fortfarande en verksamhet vid liv i januari år 2000 (under förutsättning att vi fortfarande har elektricitet förstås).
Författaren till följande dokument är Frank Hayes, reporter vid den ansedda tidningen Computerworld.
Slow start for 2000 fix
Frank Hayes
Ok, I admit it. I snickered when I heard Microsoft is finally going to announce a line of tools and services for fixing year 2000 problems -- in December.
December, I thought. December 1998. Barely a year before the corks pop, the balls drop and the computers stop. Is anybody really waiting for a product with "Microsoft" on the box to start cleaning up their millennium mess? Any company that isn't already well into a full-blown Y2K project is -- to put it bluntly -- roadkill, dead meat, somebody else's lunch.
Then I pulled up the statistics: Gartner's latest guess is that roughly a quarter of all businesses haven't even started their year 2000 fixes. And conventional wisdom says if you haven't started by now -- well, you're done for. You're out of time. You've waited too long. Microsoft or no Microsoft, it's just plain too late.
Is it? Well, it's late, all right. Probably so late that you won't be able to dodge much of the pain when the clock strikes at midnight 13 months from now. And it's almost surely too late to start a conventional year 2000 project.
But it's not too late to survive.
It's not too late to throw your whole company, not just your IT shop but every single user, at the task of inventorying every computer, every program and every database in the organization. Users know where the boxes are buried. They won't turn them all up on the first pass -- but they'll do a lot better than IT on its own.
It's not too late for the ugliest kind of triage, letting important systems fail so the absolutely critical few can be fixed. That means letting all your users know exactly what systems you won't be able to save -- and handing back to user departments the job of fixing and upgrading the hardware and software your IT shop can't.
It's not too late to replace everything you possibly can -- whether you're ripping out accounting systems and shoving SAP and Oracle applications in their place or punting thousands of iffy PCs and dropping in new hardware. Yeah, it's expensive, and users may have to completely change how they work with the new systems. But at least they'll still be working.
It's not too late to start testing everything you can't replace, testing every fix you make, testing every promise you get from vendors, new or old. And it's not too late to put users to work testing everything they touch inside your business.
Will you get it all done in time?
No. And you'll have to tell users what got fixed and what didn't, so they can work around the land mines and help spot the fixes that didn't take.
Will all your users cooperate?
Nope. Some would rather watch the company burn than do an hour of your job for you. That means you'd better sing the praises of the ones who do help out till your voice is raw.
Will things ever be the same once this is all over?
Not a chance. The users who help save your systems will never look at you the same way again. After this, they'll hold you accountable for every feature you promise, every upgrade you install, every system you deliver. And they'll never let go of the investment they made in their systems.
But at least it's not too late. You may have to turn the universe upside down -- but you're not roadkill yet.
So I've stopped snickering about Microsoft's being so late. If you haven't started your Y2K fix yet, you need all the help you can get. But it's not Microsoft or any other vendor who will save you. The only ones who can keep you from being too late are your users.
Källa: Computerworld