You've heard a lot about Windows XP's new "Service Pack 2" (SP2) lately here on InformationWeek.com, on TechWeb, and elsewhere on the Web. (See, for example, Patch as Patch Can, Microsoft Makes XP Service Pack 2 Available Via Update And CD, IE Flaw Affects Windows XP SP2 Systems and more.

I've also covered the huge new patch on my site and in my newsletter. In fact, in a recent issue I asked readers to share their first-hand experiences with SP2. After all, punditry and prognostication only go so far: At some point, it's better to hear about real-life experiences in real production environments and on real desktops.

Within days, I'd gotten hundreds of user reports, ranging from simple one-line "thumbs up/down" messages to detailed analyses of what went right and wrong. It made for fascinating reading--a wealth of hard-won, first-hand information.

At first, I started to boil down the E-mails, trying to develop statistics. But I soon realized that was unwise: Although I could perform simple counts and ratios, boiling down the E-mails lost their essence and removed the most-important information, which was in the detail.

So, I decided on a "Plan B." Yes, I'd do the rough counts and ratios, but I'd also present to you a broad sampling of the reader letters in their original form and in each writer's own voice so you could gain not only the raw "pass/fail" information on SP2, but also see exactly what kind of problems cropped up and what kinds of systems had no trouble at all.

Note that many of the reader reports I got were not 100% clear-cut. Often, a reader reporting overall satisfaction with SP2 would still cite several minor glitches and snags; or someone reporting unhappiness with the patch might still say, in essence, "everything worked fine except for this problem...."

So I categorized the reader mail into two large groups: general satisfaction, and general dissatisfaction. Where possible, I used each reader's own assessment, even if I might have judged a given problem to be more or less severe than did the reader: My thought was to let the reader's voice sound through, uncolored by me.

Although this is primarily an anecdotal (nonstatistical) exercise, it's still interesting to note that, overall, readers reported more SP2 successes than failures by about a two to one margin. But a number of the reports of successes with SP2 involved multiple machine installations (in some cases, hundreds of PCs), so the actual success/failure ratio is even higher. SP2 does seem to work, more often than not.

But there were still enough problems to suggest that SP2 should be approached with caution. While some of the reported problems are relatively minor, others are true show-stoppers that require either uninstalling SP2, or totally reinstalling XP from scratch.

So, I still recommend against full-scale roll-out of SP2 until and unless

• You've successfully completed a pilot roll-out in your environment and
• You have a full backup (preferably a disk image) of each system you'll be upgrading.

Although the overall reader experiences are encouraging--more positives than negatives--there clearly are enough rough edges to warrant great caution before opening the floodgates to SP2 in your enterprise.

But you can make up your own mind: Here's a sampling from the first 100 or so of the reader E-mails I got from early adopters of SP2. In these notes, you'll hear your peers, fellow professionals, end users--users of all stripes--say in their own words what worked and what didn't with SP2.

Later, after you've read these, please add your voice: Click over to the discussion area and tell us your SP2 experience!