Re: advpack:reg.c: Cast-qual warning fix

[Alexandre Julliard]

Andrew Talbot  writes:
> --- a/dlls/advpack/reg.c	2006-06-14 12:55:31.000000000 +0100
> +++ b/dlls/advpack/reg.c	2006-08-28 17:00:18.000000000 +0100
> @@ -255,11 +255,14 @@
>      /* FIXME: read AdvOptions val for dwFlags */
>      ZeroMemory(&cabinfo, sizeof(CABINFOW));
>      cabinfo.pszInf = tmp_ini_path;
> - cabinfo.pszSection = (LPWSTR)pszSection;
> - cabinfo.pszSection = HeapAlloc(GetProcessHeap(), 0, (lstrlenW(pszSection) + 1) * sizeof(WCHAR));
> - lstrcpyW(cabinfo.pszSection, pszSection);
> cabinfo.dwFlags = 0;

I don't think we want to duplicate strings just because of the
warning. If the string really needs to be writable sure, but if it's
just because of MS broken prototypes it's not worth the extra cost.

advpack:reg.c: Cast-qual warning fix

[Andrew Talbot]

Alexandre Julliard wrote:
> I don't think we want to duplicate strings just because of the
> warning. If the string really needs to be writable sure, but if it's
> just because of MS broken prototypes it's not worth the extra cost.

So am I right to infer that, in such cases, you would rather that I leave
the disqualifying casts in place?

Thanks,

-- Andy.

advpack:reg.c: Cast-qual warning fix

[Alexandre Julliard]

Andrew Talbot  writes:
> So am I right to infer that, in such cases, you would rather that I leave
> the disqualifying casts in place?

For the moment, yes. Let's look at this again once all the places that
can be fixed properly are done.