Vista SP1 hands-on: six trouble-free upgrades

On Wednesday, FedEx delivered DVDs containing the final, RTM bits of Windows Vista Service Pack 1. My colleagues George Ou and Adrian Kingsley-Hughes received similar packages at roughly the same time.

George was first out of the gate, reporting on his “near death” experiences with a desktop and a notebook PC. I was baffled when I read the original piece, which was completely at odds with every experience I’ve had with SP1 throughout its beta cycle and with the delivered RTM bits. The issue came into clearer focus later in the day, when George updated that original post to note that the problems he encountered had little or nothing to do with SP1. The notebook suffered a catastrophic hardware failure. And the desktop PC? As George notes much later in the story, “My desktop computer seems to be a lot healthier now after I installed Vista SP1.” George has since extensively updated the post, rewriting nearly every paragraph and acknowledging that several of his original speculations were incorrect. backtracking from almost every conclusion he drew originally. [See note at end of post.]

A few hours later, Adrian reported that his first SP1 installation “went without a hitch” and then followed up with a report that says “the promised performance gains are there.”

Meanwhile, over the past 24 hours I’ve been busy installing the RTM bits of SP1 on three different machines here, alongside three other machines that had already been upgraded via Windows Update with a release candidate that turned out to be the final build after all.

And guess what? It all just worked.

• On my main desktop system, a one-week-old Dell XPS 420 running Windows Vista Ultimate, the complete SP1 setup, start to finish, took just under 49 minutes from start (insert DVD, click OK on installer dialog box) to finish (at the Windows desktop, ready to work). It restarted three times during the installation.
• On my wife’s 18–month-old Acer notebook, also running Windows Vista Ultimate, I encountered a minor stumbling block yesterday morning when I first tried to install SP1. It stopped almost immediately, warning me that an installed language pack was incompatible the the version of SP1 I was trying to install. After some digging in Control Panel’s Regional and Language Options dialog box, I found the culprit. Last June I had installed the Italian and Spanish language packs on this machine, and the initial release of SP1 is only compatible with five languages: English, German, Japanese, French, and Spanish. (The all-language version will be ready in April.) After I removed the Italian language pack, SP1 install proceeded smoothly. I have no idea how long the actual installation took. When I came back 90 minutes later, the installation had completed and my wife was checking her e-mail and browsing the web. After a day of use, she reports no issues.
• My main Media Center PC, a year-old Dell XPS 410 that was upgraded with CableCARD support last October, just completed the upgrade. It took a grand total of 57 minutes, with no manual intervention required, and all of its Media Center functions, including communicating to a pair of Linksys extenders in other rooms, are working well.

The other three machines on which Vista SP1 is running are working properly, with no issues. For the last few months, I’ve been monitoring the private and public newsgroups carefully. Based on that feedback (thousands of hands-on reports), I know that my experience is typical. There is a known issue with some third-party drivers that don’t follow proper installation practices and can screw up an SP1 installation. That issue should affect a small percentage of Vista users, but it’s serious enough that Microsoft decided to delay the official release of SP1 by about six weeks. Aside from that issue, which will be fixed when the hardware companies release new drivers, I have seen very few SP1–related flaws.

Sadly, anyone who came to ZDNet yesterday read George’s inflammatory (and ultimately incorrect) post blaming SP1 for what actually turned out to be a hardware failure. How many people passed along those links? Quite a few, I suspect. How many people believed the incorrect information and concluded that SP1 is a buggy piece of crap? Too many. How many people came back and read George’s extensive corrections? How many people noted that the inflammatory original headline (“Death encounters with Vista SP1 RTM”) had been revised to the more prosaic and accurate “First experiences with Vista SP1 RTM”? Only a handful, I’m afraid.

I’ll be doing my benchmarking tests over the weekend and will post the results as soon as they’re ready. I’d rather be right than first.

[Update: In an e-mail to me, George points out that he tried not to draw conclusions and did indeed point out that the problems might not have been related to SP1. After rereading his post, I agree with him. The trouble is, the original, inflammatory headline and the first two paragraphs leave the unmistakable impression that SP1 caused a "near death" and a "real death" experience for two computers. Hundreds of thousands (maybe millions) of ZDNet newsletter subscribers will see only that headline and the blurb that goes with it, and they will, quite naturally, jump to conclusions. No matter how factual the complete post may have been, the short story it and the headline told was wrong.]