Re: [1/2] wininet: Partially implement FreeUrlCacheSpaceW. (try 2)

[Alexandre Julliard]

Alexander Scott-Johns  writes:
> @@ -1660,6 +1660,16 @@ BOOLAPI DeleteUrlCacheEntryA(LPCSTR);
>  BOOLAPI DeleteUrlCacheEntryW(LPCWSTR);
>  #define DeleteUrlCacheEntry  WINELIB_NAME_AW(DeleteUrlCacheEntry)
>
> +/* FCS_ flags and FreeUrlCacheSpace are no longer documented */
> +#define FCS_PERCENT_CACHE_SPACE  0  /* guessed value */
> +#define FCS_PERCENT_DISK_SPACE   1  /* guessed value */
> +#define FCS_ABSOLUTE_SIZE        2  /* guessed value */

Where do these come from then?  If you made them up they shouldn't go in
a public header.

[1/2] wininet: Partially implement FreeUrlCacheSpaceW. (try 2)

[Alexander Scott-Johns]

On 10 March 2011 14:57, Alexandre Julliard  wrote:
> Alexander Scott-Johns  writes:
> > @@ -1660,6 +1660,16 @@ BOOLAPI DeleteUrlCacheEntryA(LPCSTR);
> >  BOOLAPI DeleteUrlCacheEntryW(LPCWSTR);
> >  #define DeleteUrlCacheEntry  WINELIB_NAME_AW(DeleteUrlCacheEntry)
> >
> > +/* FCS_ flags and FreeUrlCacheSpace are no longer documented */
> > +#define FCS_PERCENT_CACHE_SPACE  0  /* guessed value */
> > +#define FCS_PERCENT_DISK_SPACE   1  /* guessed value */
> > +#define FCS_ABSOLUTE_SIZE        2  /* guessed value */
>
> Where do these come from then?  If you made them up they shouldn't go in
> a public header.

The names come from an old copy of the Internet SDK,
http://web.archive.org/web/*/activex.adsp.or.jp/japanese/specs/inetsdk.htm,
but it doesn't give the actual values. From the order they appear in
the documentation, and that native apparently uses 0 when calling
FreeUrlCacheSpace, I guessed 0 was FCS_PERCENT_CACHE_SPACE. Should I
just use a literal 0 instead?

Best regards,
Alex Scott-Johns

[1/2] wininet: Partially implement FreeUrlCacheSpaceW. (try 2)

[Alexandre Julliard]

Alexander Scott-Johns  writes:
> The names come from an old copy of the Internet SDK,
> http://web.archive.org/web/*/activex.adsp.or.jp/japanese/specs/inetsdk.htm,
> but it doesn't give the actual values. From the order they appear in
> the documentation, and that native apparently uses 0 when calling
> FreeUrlCacheSpace, I guessed 0 was FCS_PERCENT_CACHE_SPACE. Should I
> just use a literal 0 instead?

No, if these used to be in the SDK it's OK to make them public. Though a
few test cases wouldn't hurt to confirm the accepted values.