Re: DOCBOOK: Re: A proposal to clarify the semantics of DocBookgraphics

[Yann Dirson <ydirson@fr.alcove.com>]

On Thu, Feb 07, 2002 at 08:43:24AM -0500, Norman Walsh wrote:
> and added a demo at
> 
>   http://www.oasis-open.org/docbook/proposals/graphicatts/demo.html

Cool, that helps to understand what you mean.


> / Yann Dirson <ydirson@fr.alcove.com> was heard to say:
> | On Wed, Feb 06, 2002 at 10:31:32AM -0500, Norman Walsh wrote:
> |> / Yann Dirson <ydirson@fr.alcove.com> was heard to say:
> |> | On Wed, Feb 06, 2002 at 08:59:41AM -0500, Norman Walsh wrote:
> [...]
> |> If I say content-width="3in", I want the image to be 3in across.
> |
> | Hm.  I'm not sure I understand the wording "to be 3in across".  How
> | does it compare with width="3in" ?
> 
> content-width="3in" says the image (before scaling) should be 3in
> wide. width="3in" says that the area reserved for presentation of the
> image (irrespective of scaling) should be 3in wide.
> 
> | I was not talking about the respective values of "content-width" and
> | "width" attributes, but about those of the "content-width" attribute
> | and the real size of the image.  In other words, what is supposed to
> | happen if you lie to the processor when you specify "content-width".
> 
> The content is scaled.

What bothers me then, is that the result here will be likely to be
different if the processor can't determine the image size.  But I can
now see how this is useful.

Maybe we would need an additional pair of attributes to specify the
"real size" (physicalwidth/physicalheight ?) of the image, so that we
can get still the benefit of the width/height behaviour you describe
even when the processor cannot determine it itself.  Here we could
state that telling lies leads to implementing-defined behaviour.

-- 
Yann Dirson <Yann.Dirson@fr.alcove.com>                 http://www.alcove.com/
Free-Software Engineer				      Ingénieur Logiciel-Libre
Free-Software time manager    	       Responsable du temps Informatique-Libre
Debian GNU/Linux developper <dirson@debian.org>