Re: Regression Test - kernel atom test

[Alexandre Julliard]

"Dmitry Timoshkov"  writes:
> - MS eventually will drop support for NT4 someday (as they did for Win95)
>
> and it's better IMO to have it mind now than try to catch up later.

Sure, once applications start depending on Win2000 features/bugs we'll
have to implement them. But we don't know yet which of these features
will be important; and implementing every single one of them is a huge
waste of time since the vast majority of them will never matter.
> - What to do with GlobalGetAtomName on short buffer and GlobalDeleteAtom
>
> tests which behave differently under different Windows implementations?
> Drop them altogether?

Yes. If a test does not behave consistently, it means it doesn't test
an exported feature of the API, but just a random side effect. These
tests are useless and should be dropped, unless we identify an app
that relies on the side effect.