W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > January to March 2005

Re: A note about the definition of "structure"

From: Michele Diodati <michele.diodati@gmail.com>
Date: Wed, 19 Jan 2005 19:06:46 +0100
Message-ID: <2e1e87c050119100667b19fe2@mail.gmail.com>
To: w3c-wai-gl@w3.org

Hi Gregg,

you wrote:

> I don't follow you.
> In a hierarchical relationship the parent cannot be a child of the child (or
> a grandchild). In a situation that is a web you will end up with this
> situation - and that doesn't match any definition of hierarchy that I know
> of.
> Am I missing something?

Please, can you show me an example of a relationship in which the
parent ends up being the child of the child?

In my opinion, the hierarchical viewpoint is only one of many possible
viewpoints about structural relationships between objects in the real
world. Hierarchical relationships tell us about the status of
inclusion or exclusion of a given object, or category of objects, with
respect a given feature, category, object, container, etc. On the
strenght of hierarchical relationships, we may know if two object are
siblings in a given hierarchy or if one of the two is an ancestor (or
a descendant) of the other, or vice versa. As I said in a previous
message, the presence of a visible hierarchical relationship between
two objects, as two different chapters of the same book, depends on
the level of abstraction (depth, universality) at which we consider
the relationship between the two objects. For example, between two
brothers there is an evident horizontal relationship: both are sons of
the same father; on the contrary, between two not related human beings
it seems that no hierarchical (horizontal or vertical) relationship
exist. But, if we enlarge our viewpoint, and consider them for example
as inhabitants of the same town, we can put them in a hierarchical
relationship. It' a viewpoint issue. Depending upon our viewpoint, two
object can be the former child of the latter under a certain respect,
and the latter child of the former under a different respect.
According to me, it may resolve the ambiguity you report.

However, it seems to me that the core of the problem is not whether
non-hierarchical relationships exist or not. The core of the problem
is to eliminate from WCAG 2.0 any ambiguity about the context in which
the word "structure" is used. GL 1.3 says "Ensure that information,
functionality, and structure are separable from presentation." Here I
understand "structure" as a question of markup: code your web page in
such a manner that you can easily split structural code from
presentational code. On the contrary, the definition of "structure" in
the glossary of the same document says: "A book is divided into
chapters, paragraphs, lists, etc (...) A bicycle is divided into
wheels and a frame (...)." Here "structure" is the series of
relationships existing between the parts of objects belonging to the
real world. It is a great contradiction, and harmful for the
comprehensibility and applicability of the spec. In the real world
(printed, spoken, on screen documents) structure is not really
separable from presentation. Instead, at the code level, there is no
problem splitting structural code from presentational code.

I know it is a bit difficult, but I don't know how to explain the
problem in an easier way.

Received on Wednesday, 19 January 2005 18:07:19 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 16 February 2017 16:48:57 UTC