On Tuesday 12 February 2008, James Hawkins wrote:
> On Feb 12, 2008 8:16 PM, Peter Dons Tychsen <donpedro@tdcadsl.dk> wrote:
> > Hello M.
> >
> > What is wrong with detecting the version, and branching the test code
> > accordingly? Other tests do that IIRC. I think that is a better
> > solution, if the version selector in winecfg is still going to be
> > meaningful (registry).
>
> The only time the tests should check for the windows version is if we
> know an app does the same thing, which is hardly ever.

Well, if we have an old app from win9x times, it will expect windows 95 
behavior, without checking for windows version of course.

So if Wine is set to impersonate win95, then it should behave like on win95 
and the tests should check windows (here: wine) version to select a proper 
test. Such a version-specific test should pass on win95 and wine with win95 
personality chosen. It should also pass on other versions, by choosing a 
different path (and possibly doing nothing if it doesn't apply to that 
windows version).

So, for example, methinks all applicable unicode tests should be disabled on 
the early win95 versions which had no such support (if so bolt-on unicode is 
detected).

Wouldn't that make sense?

Cheers, Kuba