There’s been a furious counterpush by some over the claim that Vista needs to be largely rewritten, with some Microsoft employees demanding on their blogs that anybody reporting such things be fired (we’ll talk about this guy some more tomorrow).
Well, there’s two ways one can take that story. One can take it absolutely literally, or a bit more figuratively, like this very informative assessment.
If you take it absolutely literally, no, it’s not likely that MS will rewrite 30 million lines of code in the next ten (not two) months.
However, it is quite easy to believe that sections of Windows amounting to 30 million lines of code are due for a second look, or that, in the estimation of someone, sections of Windows amounting to 30 million lines of code ought to be due for a second draft, because they don’t quite work right or are hugely bloated.
That’s pretty much what a marketing director of Acer had to say about it.
The Hole In the Roof
Let’s say your name is Ted. One day, you go on your roof, and you notice your neighbor’s roof has a big hole in it. You tell the other neighbors, “Joe has a three foot hole in his roof, he’s going to have to bring a roofing guy in.”
Joe hears this and gets greatly. He goes up and measures the hole, and finds out it’s only two-and-a-half feet wide. He tells the neighbors, “Ted is a liar. There is no three foot hole in my roof.”
Ted may not have quite accurate, but did he really lie? Isn’t the hole still there, and isn’t it bad?
Let’s say Joe decided a long time ago not to fix the roof. He tells his neighbors “Ted is a liar. I find no need to get my roof fixed.”
Again, Ted didn’t get it right, but did he really lie? Isn’t the hole still there, and isn’t it bad?
Finally, let’s say Joe put a bathtub under the hole in the roof, and tells his neighbors, “Ted is a liar. That hole is an improvement; it’s there to supply my indoor swimming pool.”
You know the routine.
What do we know about Vista, as it is today? As it stands right now, the thing’s a pig. P-I-G, pig. This isn’t some anonymous tip, Microsoft publicly said so. The way it looks now, Joe Sixpack is going to need between a half to a full gigabyte of RAM, plus a lot more elsewhere, to browse and write emails comfortably? Doesn’t that indicate something seriously wrong?
If you have a bunch of code that runs as slow as molasses, and you decide not to fix it, does that mean there’s no problem?
If you have a bunch of code that just doesn’t work right, and you drop the feature, did you really solve the problem?
If your OS chews up more memory than Ken Jennings uses to play Jeopardy just to play Solitaire, and you say, “It’s not a problem, it’s a feature,” does that make everything OK?
If you’ve ripped out most of your initial goals, and six months before release, you still have a dysfunctional resource pig, should you be congratulated for delaying the project?
The issue here is not exactly how many millions of lines of code are going to be rewritten, or should be rewritten. The issue is how badly the millions of lines of code there now function, either slowly, inconsistently, or not at all.
The issue here is not whether it is a wise decision to delay shipping a POS a few extra months. The issue is “Why was it such a POS at this late stage of the game?”
In short, the issue is not the exact details of a report of wrongness, but the current degree of that wrongness.
There are people blogging at Microsoft saying the 60% statement is erroneous. OK, if the information is bad, then you can give us better. Tell us why this is being delayed, too.
There are people blogging at Microsoft saying things work fine now. OK, if it works so great, why isn’t it being released on time? If it works so great, why does it need four times the memory of XP to do the same thing?
The issue here is not about the wisdom of a particular delay. The issue is about whether MS is losing the ability to function.
As Paul Thurrott, who certainly isn’t an MS hater, reminded his readers recently:
. . . Vista is crashing down around them after years of delays. This product was originally scheduled for release in 2003. Microsoft co-president Jim Allchin, who currently runs the Platform Products and Services division with Johnson, has already taken the biggest step in turning around the division. In 2004, he effectively shut down Vista development and started over from scratch after recognizing that the project was going nowhere fast.
So Vista has been a sick puppy for a long, long time.
Thurrott also said why the puppy is so sick:
“So Microsoft is finally shaking up the Windows Division and putting Office guru Steven Sinofsky in charge. Great. My advice to Sinofsky is simple: Shake the entire tree and get rid of the dead wood. Re-examine everyone involved with Windows, from current employees all the way down to beta testers and MVPs, and ensure that these people are valuable, productive, and providing honest and timely feedback. This is the time to make the Windows division a lean and mean fighting machine, one that can rise to the challenges imposes by Apple’s Mac OS X and Linux. History provides plenty of examples of what happens to former market leaders that get fat and lazy. Sadly, the Windows Division is one of them.”