We’ve got a projector in the office which gathers various streams of news and shows them on the wall, and today it keeps talking about Windows 7. It seems SteveB is making a speech tonight and everyone expects it to announce the beta of Windows 7. (Mary Jo has some more ideas what he might say)
So in preparation for its arrival here are a handful of thoughts about beta testing , and the next version of Windows.
1. Remember what a beta is for. It’s a two way thing; you discover what might have problems, what’s new and great and what’s new that you just don’t like. You test thoroughly. Try those crappy old apps and old bits of hardware (I’m told that some things which need coaxing to work on Vista are more likely to work on 7 out of the box. If it can’t be made to work on Vista with the app compat toolkit, it probably can’t be made to work on 7 either). It might also be your first tilt at IE 8. But it is a two way process: if you find something which doesn’t work we want you to report it. That was our reason for letting you have it.
2. Products ship when they are ready part 1. I’ve seen all kinds of rumours about when Windows 7 client and Server 2008 R2 will ship. One intriguing one is that PCs shipped after July 1 will get a free upgrade. Free upgrades from RTM onwards used to be the rule. With Vista we had upgrades 24th October 2006, and RTM was in November. If this rumour is turns out to be true that says a release will not be much after July 1. I’ve always reckoned on 3-4 months for a beta and a month for a release candidate as a good rule of thumb. There isn’t time to do usual two betas and two or three release candidates by July, which makes another rumour – of only one beta the only way that will work. I don’t have any inside scoop on this. We said the new client OS be 3 years after Vista. Exactly 3 years means RTM in November and launch parties in 2010 - nice fit for my timescales for 2 betas and 3 RCs . Unless something is said in the Ballmer speech remember my old saying, those who really know don’t talk and those who talk don’t really know. Some great things are already being said about 7, but
3. Products ship when they are ready, part 2. A beta, by definition is not ready. Life will not be free of all disruption I’d be surprised if anything in the a beta OS trashes my data. But staring at a dead file system with nothing to do but mutter “that was a surprise” * isn’t something I’m going to let happen. So I’m going to try to get a fresh hard-disk and copy my data to it and leave the old one alone. And I’m going to be sure to test the backup and restore system :-)
4. A lot will be written about the OS, much will be junk . Jason Perlow at ZDNET gives a great example. “There’s no run menu”. Press the [Window] key instead of [Window] & [R] or click start (rather than click start, click run) and type what you would have typed in the Run box. It works the same, and its quicker. And it finds things before you’ve typed the whole name. The mentality of saying “I must have my run box in floating window named run and it can’t be merged with search” is just… well, Jason’s colleague Ed Bott just stops short of calling him a luddite. There are other good points in Ed’s post too. I’ve long held that he knows what he’s talking about.
* Back in my days in Microsoft consulting services , we would be asked to review designs and tell the customer that they good be guaranteed to work. I’d always explain that such guarantees are impossible but we can say we’ve reviewed it and if competently implemented nothing about it leads us to expect a problem. Of course if a problem arises we will say “COR ! That was unexpected”. That always raised a laugh, but the serious point was we