ARGB cursors (was Re: Adjusting the 2.4 schedule)

[Owen Taylor <otaylor redhat com>]

On Tue, 2003-06-24 at 19:06, Matthias Clasen wrote:

> - There is a new patch in bugzilla for bug #69436 (ARGB cursors)
> 
> - The issues brought up last time were
> 
>   - Coding style: hopefully fixed.
> 
>   - X specific docs: 
>     I tried separate each doc comment into a generic paragraph and 
>     an X specific one.                  

Looks good, though it might be good to point to the
Query functions.

>   - Might be better to use a frame struct instead of parallel arrays:
>     I introduced GdkCursorImage for that purpose

Looks good.

>   - Might be nice to make the capabilities of the system queryable: 
>     I added four gdk_display_*() functions for this purpose
>     which closely mirror the corresponding Xcursor interfaces.

Hmm, I think an important thing here which is missing is the *maximum*
allowed cursor size. Though this can be a fuzzy concept:

 Windows 98 and earlier have a fixed maximum size that can be
  queried from the system.
 Classic X cursors have XQueryBestCursor(), an undocumented and
  unclear function. The XLib docs say:
        
Some displays allow larger cursors than other displays.  The  
XQueryBestCursor function provides a way to find out what size cursors 
are actually possible on the display.  It returns the largest size that 
can be displayed.  Applications should be prepared to use smaller
cursors on displays that cannot support large ones.

 But the exact interpretation is unclear. For XFree86,
 XQueryBestCursor() will clamp it's returned result to the maximum
 allowed hardware size, but use software fallbacks for larger cursors.
 I think other servers may deal with over-large cursors in less-graceful
 ways.

I think a good use case for the query API is:

 Could gtkdnd.c use cursors instead of icons to display the DND
 drag icon? (Ignoring the problem of overlaying two cursors into
 a single cursor)

> - Are we ready to commit? / I'm going to go ahead and 
>   commit in a week if nobody objects. [*]
> 
> Matthias
> 
> [*] What was that footnote supposed to contain, Owen :-)

I'm not sure at this point; I'll let people imagine something
interesting for themselves.

Regards,
					Owen