"Alex Brown" <> wrote on 04/24/2009 10:27:46 AM:
> Rob hi
>
> > The concept of soft
> > versus harsh light has been around since at least Caravaggio
> > (1571-1610). This is not a term specific to computer
> > graphics. It is a general term referring to scene lighting
> > and we use it in its ordinary sense.
>
> Just because a term has been around for a long time and is thought to be
> well-understood in an "ordinary sense", does not make it suitable for a
> technical specification - otherwise we'd be okay describing things
> shapes as "good", "graceful", "beautiful" etc.
>
I disagree. You need to look at the scope of the standard and what
expectations for interoperability for a particular feature. ODF is not a
PDF. ODF is not JPG. The scope does not call for pixel perfection. We're
representing documents at a higher level, fully aware that there will be
some rendering differences in the details. We are not obligated to define
exactly what colorimetric value is required when we say something is
"red". We are not obligated to specify the exact radius of curvature when
we say a text box has "rounded edges". And we do not need to specify the
exact size, in steradians, of a harsh light source.
The question is whether the text specifies requirements to a degree that
the desired balance has been met, between having enough specificity to
allow implementors to meet interoperability expectations of users, without
over-specifying to the degree that it raises implementation costs without
appreciable user benefit. And remember, over-specifying can reduce
interoperability as well, by tying a feature closely to a particular
rendering environment.
The question of exactly what the desired balance should be is a question
for the ODF TC to decide when we review your comment.
-Rob
