W3C home > Mailing lists > Public > www-style@w3.org > January 1999

Re: Cursors.......New Proposals

From: L. David Baron <dbaron@fas.harvard.edu>
Date: Tue, 12 Jan 1999 15:43:21 -0500 (EST)
Message-Id: <199901122043.PAA13407@login4.fas.harvard.edu>
To: www-style@w3.org
>From: "Albert Lui" <al@primeconsult.com>
>Date: Mon, 4 Jan 1999 23:27:48 -0500
>Subject: Cursors.......New Proposals
>
>1. NEW PROPERTY DESCRIBING COLOR OF CURSORS
>-------------------------------------------
>
>A new property called cursor-color should be adopted to allow authors 
>to specify various colors through existing color names or hex 
>numbers. This issue has not yet been addressed in any of the 
>previous CSS specifications to date, but would clearly extend the 
>capability of the author to control the presentational outlay of the 
>document.

I think this is probematic because cursors are images, and have
multiple colors.  (In Windows, white, black, transparent, and
reverse-transparent) A cursor color property would not specify which
color to change.  I think the current image formats for cursors aren't
really suited to this type of manipulation.  If/when SVG cursors come
out, that may be a different matter.

>3. "POSITIONING" CURSORS
>------------------------
>
>Currently, cursor presentation can only be rendered through an 
>element (such as TABLE). There is no mechanism for describing which 
>icon/cursor should be displayed on particular areas of the page which 
>cannot be described through the HTML coding.
>
>In addition to the advantages offered through absolute positioning, 
>for example a document in which an image acts as the background. The 
>image consists of a thick 1 inch yellow line running down the left 
>hand side of the page with white space consisting of the rest. 
>Different cursors could act for different areas of the "image"
>
>Real example: CNET's Download Centre at http://www.download.com 
>(though it has two thick lines)

What one would want here is not a cursor change over the background,
but a cursor change over the links.  There is almost definitely some
element in there that would be appropriate, probably a TD element, that
covers exactly what would be wanted.

>A simple example would be:
>
>BODY { cursor-area-top: 5px;
>       cursor-area-left: 10px;
>       cursor-area-height: 2cm;
>       cursor-area-width: 4cm;
>     }

I think this would be better left to:

BODY	{
	position: relative;
	}

DIV	{
	position: absolute;
	top: 5px;
	left: 10px;
	height: 2cm;
	width: 4cm;
	cursor: ... ;
	}

I think the idea of specifying properties on elements is a good one,
and I don't see a reason to give it up.  Normally one would want to
specify a cursor change over an already existing element, anyway.  If
not, one could create an appropriate element.


>4. DISCUSSION
>-------------
>
>There are several remaining issues which need to be dealt with:
>
>a) How would one describe the color of a predefined cursor in the 
>following example from 18.1 in which a mechanism for using a primary 
>cursor is already defined?
>
>P { cursor: url("mything.cur"), url("second.csr"), text; }
>
>b) Predefined cursor files may consist of various colors. Should the 
>stylesheet have the authority to overide the existing color scheme?
>

I think these are both serious problems.  In the interests of backward
compatibility, I think the cursor property should stay as it is now.  A
cursor-color property could be appropriate.

However, I think the second problem you mention is an essentially fatal
flaw in the proposal, as long as cursors are specified is raster image
data.  I tend to think that manipulation of colors within an image should
be beyond the scope of CSS.  Colors, if desired (and possible), could be
specified in the cursor image files themselves.

Furthermore, some of these proposals may be very difficult to
implement.  I don't think I have ever seen a colored cursor in Windows,
and I don't know if it can be done.  I haven't seen any on Mac or Unix
either, but I am less familiar with those platforms (especially Mac).
I think the CSS specs should stay within the realm of what browsers can
be reasonably expected to implement.

David Baron

--------------------------------------------------------------------
L. David Baron         | Freshman, Harvard
dbaron@fas.harvard.edu | < http://www.fas.harvard.edu/~dbaron/ >
--------------------------------------------------------------------
Received on Tuesday, 12 January 1999 15:43:22 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:53:58 GMT