Yes, Software Projects Miss Deadlines. Ask Microsoft.

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

The following is from an article in today's Wall Street Journal Page B6. I would post a link, but WSJ is a subscription website. Linkmeister, could you be of assistance somehow?

Windows 2000, formerly known as Windows NT 5, set the official date of February 17 for the program's release.

"Windows 2000 was already more than a year late before the latest announcement, suggesting that the task of bringing the giant product to market was far more difficult than Microsoft originally anticipated."

Composed of 30 Million Lines of Code.

How many lines of code are the IRS, GM, Citigroup remediating?

Late, late, and more late....

For Newbies: DON'T BE LATE ON YOUR PREPS!!!!!!!

-- PJC (paulchri@msn.com), October 27, 1999

Answers

Microsoft is supposedly the best software company in the world, having the pick of the best programmers and developers. How will the leeser-lights fare? I know. No one knows what'll happen..... What a crock-o-crapola.

-- PJC (paulchri@msn.com), October 27, 1999.

That is not true, didn't you see my show? Those little bugs happen every day and they never cause any problems. Just have your checkbook ready, and enjoy the new millenium!!

-- Bob Cringely (iam@ pieceofshit.yuppie), October 27, 1999.

Microsucks MAY have good/excellent programmers, but the bottom line is marketing, not quality. That is, it's a MANAGEMENT problem. MANAGEMENT is what causes software to be released with major bugs. And MANAGEMENT is the cause of fiascos like Y2K.

-- A (A@AisA.com), October 27, 1999.

Microsoft is supposedly the best software company in the world, having the pick of the best programmers and developers.

Bullsh*t!! Microsoft's code is CRAP! They take kids fresh out of school with little or no REAL WORLD programming experience and turn them loose to write 'cool' code, quality be damned. The Windows API is so buggy and friggin' inconsisent that I've personally wasted WEEKS trying to get a straight answer from them about the way some piece of code is *supposed* to work, or alternately, find a workaround for a bug myself. And don't EVEN get me started on their documentation (or lack thereof)....Grrr!@

-TECH32-

-- TECH32 (TECH32@NOMAIL.COM), October 27, 1999.


Yes, I know it's politically correct to hate Microsoft. And they do tend to bite off more than they can chew. Still, there are two points to bear in mind here:

1) Writing a 30 million line operating system isn't quite the same sort of task as finding and fixing date bugs in an application. You must *write* 30 million lines and get them to work, not simply fix a small minority of existing lines that already work, but won't work correctly later.

2) Even with all the admitted Microsoft bugs, millions of computers are fully operational and functioning with this code. It's not a question of whether the code is buggy or poorly documented or the API is inconsistent or the code is slow and bloated. When it comes to y2k, we're looking at *usable in a pinch*, not *perfect*. I'm not defending all the rebooting practice we get whenever we see the Blue Screen of Death. The question is, since Citigroup *must* operate with wherever the code stands come the witching hour, can they do so at all? How well?

-- Flint (flintc@mindspring.com), October 27, 1999.



With a new system, you (should!) have some blueprint of what it should do. With an existing application, you may not have documentation, so "fixing" it can be far more challenging than writing the same number of lines of new code. All of which assumes that you even have the source code...

Got one like that on the corner of my desk at this moment... It suddenly became "non-mission critical!"

-- Mad Monk (madmonk@hawaiian.net), October 27, 1999.


This is no problem. Y2k will be a non-event. The government and the papers say so. And we pollies believe everything we are told.

-- Dr. PollyDork (DrPollyDork@moron.com), October 27, 1999.

Flint,

1) Writing a 30 million line operating system isn't quite the same sort of task as finding and fixing date bugs in an application. You must *write* 30 million lines and get them to work, not simply fix a small minority of existing lines that already work, but won't work correctly later.

Yeah but they're not starting from scratch here. They're just building off of the existing NT code base which has been around for quite a while (the same can't be said for the programmers who originally wrote it though).

2) Even with all the admitted Microsoft bugs, millions of computers are fully operational and functioning with this code.

Well, that's debatable (at least you admit to not defending the reboots, etc). I think it's more a question of *HAVING* to use what you got (or had shoved down your throat ala anti-trust violations). I swear, I wish companies would start sending MS bills for the millions of wasted hours MS has cost them. If you purchased trucks from a manufacturer and those trucks broke down three times a day would you say "gee, well they *eventualy* get there...."? Hell no. You'd take'em to court to recoup your lost revenue. MS has been VERY lucky on that front. So far.

...we're looking at *usable in a pinch*, not *perfect*.....The question is, since Citigroup *must* operate with wherever the code stands come the witching hour, can they do so at all? How well?

Wrong line of thinking Flint. 'Usable in a pinch' is what got us into this whole Y2K mess in the first place, remember? Re: Citigroup - your guess is as good as mine. I know a woman who worked for them doing remediation back in 1998. She thought things were going well back then but that they had *very* tight deadlines. Know what tight deadlines mean Flint? Corners get cut, yet again.

-TECH32-

-- TECH32 (TECH32@NOMAIL.COM), October 28, 1999.


Tech32,

Actually Windows 2000 is re-written for the most part. Release Canidate 2 is already more stable then NT4 SP5. You will be pleasently surprised. I am running it for a while now and haven't run across to many bugs yet.

They took care of more then 15,000 bugs between RC1 and RC2.

Plug and Pray actually works now...

Give them a chance. Reason this has been delayed this much is that they tried to do it "right" this time. Have they succeeded...??? Don't know, but it looks good so far...

And NO, I don't get paid by Microsoft. Just like what they have done this time... Versus the piece of crap called 98.

-- STFrancis (STFrancis@heaven.com), October 28, 1999.


STFrancis,

Forgive me if I'm skeptical about the stability of a new/revised OS. Every time it's the same story "We really, really do have it fixed this time. Plug-and-Play really does work, we promise." That's what I heard about Win98 (which is MUCH WORSE than Win95) and what I was told about Win95 in relation to Win3.1. I even heard from beta-testers that they thought Win98 was the cat's meow. Crap!!

I've used Windows since Version 1.0 (you remember, back then it only had tiled windows) then Win286, Win3.0, WFWG, Win95, WinNT and lastly Win98. Every single one has been a nightmare to develop under and even worse to support. Microsoft just don't care about shipping code that they KNOW has thousands of bugs. They KNOW people will have no choice but to start using it because they twist the arms (to put it mildly) of the PC manufactures to get it installed on all new shipping PCs. This is why they are under fire (and deservedly so) for anti-trust violations.

To be fair NT 4.x is pretty good IMHO in regards to it's stability but has it's own support/security/development issues that'll drive you nuts. I doubt they really re-wrote most of it's code, more likely that they took the existing core code and are re-writing the event handling and interface stuff. If it really IS a complete (or near complete) re-write then that's even MORE reason to doubt it's stability. Microsoft's track record is horrible and they're going to have to REALLY, REALLY PROVE IT to me this time.

-TECH32-

-- TECH32 (TECH32@NOMAIL.COM), October 28, 1999.



Moderation questions? read the FAQ