Beta 1 is not safe. But it is reasonable.

In the Windows 7 beta newsgroup, a certain Mike asked:

MikeXXXXXXX wrote:
> How long should people wait until we start testing say, internal house tools
> against Windows 7? Is it safe to assume that if we were to modify our
> applications to work with Beta 1, that they would work with the final
> bug-free?

My response:

It depends.

My experience is, the more of a straight ahead application it is, the
more it does stuff in standard ways and not in some weird legacy
freakish kludgey way, the less likely it's going to be to break.
Lower level stuff (vpn clients, antivirus, various sorts of drivers)
historically seems a more fragile from build to build.

If you're asking safe in terms of "my boss will frown mightily at me
if I waste dozens of people's time, and I'll be first on the list for
the layoffs that are surely coming next quarter" - then NO it isn't
safe to modify your apps for Beta 1.

But if you would get some huge benefit out of having those internal
house tools ready for use on Win7 ASAP, and it wouldn't cost you an
unreasonable amount if you did have to do it over, it's probably
reasonable to go for it.

It all depends on what sorts of risks you're willing to tolerate, and
what kinds of benefits you stand to gain.

And a follow-up:

Do your apps work with Vista? I'd say if an app works with Vista but doesn't work with Windows7, that's a little dicier than if the app doesn't work on either.

That was a little convoluted. If the app works on Vista, and is broken on Win7, then it was something that changed back in Vista that you need to fix for. That code is probably going to remain consistent through the Win7 beta. (I'm just speaking in terms of likelihood and rules of thumb -- this isn't a scientific fact.)

But if it was something that worked in Vista and is only breaking now in Win7, then you're definitely working with some code that's changing. It might be just done with changing, but it was recently in motion and probably hasn't settled down yet.