W3C home > Mailing lists > Public > www-svg@w3.org > July 2011

Re: Behaviour of empty SVGTitleElements

From: Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>
Date: Sat, 2 Jul 2011 13:48:03 +0200
To: www-svg@w3.org, klaus.foerster@uibk.ac.at
Message-Id: <201107021348.03583.Dr.O.Hoffmann@gmx.de>
Hello,

my own understanding/interpretation of title and desc in SVG:

SVG tiny 1.2 has currently the most advanced description for
elements like these.
About behaviour of viewers:
http://www.w3.org/TR/SVGMobile12/struct.html#uiTitleDescBehavior

As you can find in the SVG tiny 1.2 recommendation, tooltip is 
not the intented presentation of a title element, if there is no
role="tooltip" provided. Other behaviour of viewers can be
interpreted as poor or wrong implementations or misunderstandings
due to less helpful descriptions in previous recommendations.
Technically however the behaviour is independent from the content
of the title, therefore there is no difference between some text or
nothing. 

And an empty title element is no meaningful use of a title
element, if the parent elements has some more content at
all (with some relevant meaning), because the title element is
a short alternative text representation of the parent element.
Empty title means simply, that the parent means nothing.
No title element is different, this provides simply no specific
text information about the parent.

There is the mentioned exception of an additional RDFa attribute 
indicating  a tooltip interpretation as intention of the author - 
but one has to keep in mind, that the content of a title element 
is always a title and no directive or suggestion what to do - 
therefore something like 'press to start' would be bad content anyway, 
better would be 'Button' or 'Button to start animation'.

If you need to remove the meaning, it is maybe the best and
meaningful approach to remove the complete parent element
to get the intented effect to remove the meaning completely ;o)

Both title and desc represent text alternatives, but there is no
behaviour defined, how to present them - this can depend on the
capability of the viewer, therefore some popup-window on demand
can be ok of course (not just hovering an element, this can be quite 
annoying for authors, wanting to have no text alternative, if the graphical
presentation is presented, but wanting to provide a text alternative for
those, who cannot see the graphical representation) , if possible, 
but this does not mean, that the title or desc becomes the meaning 
of a tooltip.
Of course a structured text alternative for the complete document
or a fragment available on demand would be a good presentation
of such elements like title and desc. Finally all desc and title applicable
for an element should be accessible somehow, maybe the innermost
first, but of course others are applicable as well.
To keep it simple, I think, such a text alternative should not be
time dependent, instead it should describe the time dependent
meaning of functionality of such a structure. This fits better to
a text alternative of complex content and covers more user
groups with quite different reasons, why they want to use the text 
alternative and not the graphical alternative of the file.

With "If user agents need to choose among multiple 
‘desc’ or ‘title’ elements for processing (e.g., to decide which string 
to use for a tooltip), the user agent shall choose the first one."
it is only discussed, if the author ignores the recommendation to
use at most one title or desc per element. For example for the
document title the viewer has to decide which one to use -
there can be only one title for the complete document.
This is quite different from the normal case, that parent elements
up to the root svg element have title and desc too.
In such a case, of course, all are somehow applicable, not just
one of them - therefore a structured presentation is necessary to
indicate, how title and desc elements are nested.

SVG 1.1 has no specific elements for tooltips, but one can do this
of course with interactive and declarative animation. And one can
use the element metadata + RDF to indicate the intented meaning.
SVG tiny 1.2 has RDFa attributes available and one can use them
to indicate any element to be a tooltip. 
My suggestion for SVG tiny 1.2  would be to use the metadata element
with role="tooltip" and the value of content as the tooltip text - what does
not mean, that many current viewer will manage to present a tooltip indicated
as an RDFa tooltip today. But this may change, if more authors call for
tooltip functionalities.
If the tooltip text needs to be changed, interactive declarative animation
should be typically a good solution as well, still one can indicate the
complete construction as a tooltip and describe the complete behaviour
in a desc element of such a construction to keep it accessible.

Of course, a specific tooltip element in SVG 2 would be quite useful,
to keep it simple for authors to provide simple  tooltips in SVG without
confusing the meaning of other elements by abusing them for such
a functionality and without the requirement for interactive declarative
animation for such a functionality.

Olaf
Received on Saturday, 2 July 2011 11:48:44 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:48 GMT