Re: Realtime/embedded project to help with employment.

["Robert I. Eachus" <rieachus@attbi.com>]

Alexander Kopilovitch wrote:

> But I don't see other way for Microsoft, given that it has tremendous quantity
> of unskilled users, many third parties involved, and at the same time it
> reasonably wants to restrict its responsibility, at least separate itself from
> those third parties. How can Microsoft test an update if it do not replace
> foreign drivers? Then, if the update installer will ask user, this will mean
> that responsibility for subsequent failure (incompatibility) will be somehow
> divided. While with current mode of operation the issue of responsibility is
> much cleaner - if Windows works properly after update, but fails after subsequent
> re-installation of some third-party driver then it is easy to blame the driver
> and to defend this position.

Which assumes that Windows will work properly after the update.  But the 
reality is that the third party vendor had a reason for distributing an 
update to their driver.  Reinstalling the earlier version that is part 
of Windows is going to break something, or reinstall a security hole 
that the user thought he had fixed.

IF Microsoft asked before replacing a third-party driver, and had all 
sorts of warnings about why you might have problems if you don't that 
would be fine.  And of course that is the DEFAULT behaviour of the 
Microsoft supplied installer when used with other company's 
applications.  AFAIK, only Microsoft overrides it so there is neither a 
warning or a record of what was done.

-- 
"As far as I'm concerned, war always means failure." -- Jacques Chirac, 
President of France
"As far as France is concerned, you're right." -- Rush Limbaugh