Re: [try2] kernel32: Add partial stub for NeedCurrentDirectoryForExePath - what's wrong?

[Kirill K. Smirnov]

> > For containing backshash. Yes, comment must be better.
>
> A comment that states you're looking for a backslash is superfluous;
> the code makes that obvious.  A better comment would be 'check for
> filename only' or something along those lines.
> Maybe, 'MSDN claims that if filename contains backshash then return value is 
> always TRUE'?
> How do you know that's 
> what this function does if you haven't tested it?
> I've mentioned an MSDN article in the patch. There are some glimpses what this 
> function does. Easy, isn't it? ;-)

That MSDN article is rather strange: it claims, that filename is the name of 
executable file, but in wine traces of cmd.exe I can see '.' (dot) as an 
argument.

--
Kirill

[try2] kernel32: Add partial stub for NeedCurrentDirectoryForExePath - what's wrong?

[James Hawkins]

On 4/25/07, Kirill K. Smirnov <lich(a)math.spbu.ru> wrote:
> > > For containing backshash. Yes, comment must be better.
> >
> > A comment that states you're looking for a backslash is superfluous;
> > the code makes that obvious.  A better comment would be 'check for
> > filename only' or something along those lines.
> > Maybe, 'MSDN claims that if filename contains backshash then return value is
> > always TRUE'?

Just submit some test cases and you won't have to say anything about
msdn.  We match the implementation of Windows, not msdn (though most
times they're the same, that's not always the case).
> > How do you know that's
> > what this function does if you haven't tested it?
> > I've mentioned an MSDN article in the patch. There are some glimpses what this
> > function does. Easy, isn't it? ;-)
>
> That MSDN article is rather strange: it claims, that filename is the name of
> executable file, but in wine traces of cmd.exe I can see '.' (dot) as an
> argument.

A test case would easily clear this up.

-- 
James Hawkins