Lists Home |
Date Index |
Hypertext Reference: semantics depend on application.
Use the same name for many different semantics,
and yes, it becomes meaningless. See Boltzman entropy.
From: Micah Dubinko [mailto:MDubinko@cardiff.com]
Does anyone know of any design notes on the Web going all the way back to
In particular, I'm examining the attribute 'href'. Is there a definitive
statement of what the 'h' stands for? What are the communal expectations for
something named 'href'?
Attempting to reverse engineer the HTML spec (at the 4.01 version), it
appears that the attribute 'href' applies to user-activated,
content-replacing links. (Other uses of URIs use differently named
attributes, like 'src', 'cite', 'profile', etc.) <base> and <link> also seem
to be exceptions.
Rumors are that XHTML 2.0 has a greatly expanded use of 'href' as a
XInclude uses 'href' in an automated sense. Is that an appropriate use of
the name? Similarly, XForms uses 'href' to automatically fetch things. Is
that violating some design principle?
The big design question is: when is 'href' the right name for an attribute?
Has is become a meaningless name?