W3C home > Mailing lists > Public > www-html-editor@w3.org > April to June 2003

comments about XHTML 2 WD nr. 4

From: Rafael Gieschke <rafael@gieschke.de>
Date: Mon, 28 Apr 2003 20:25:21 +0200
Message-ID: <00ce01c30db3$995e3160$bac906d5@computer>
To: <www-html-editor@w3.org>

Hello dear all,

I have "some" comments about XHTML 2 WD nr. 4 (xhtml2-20030131):

5.5. Attribute Types

"Authors may use the following recognized link types [...]"
May authors also use other, not mentionend, link types?

LinkTypes: Alternate

There is no more an attribute called hreflang ("When used together with the
hreflang attribute [...]"). This should be replaced with "xml:lang".

LinkTypes: Start, Next, Prev

The description of "Start" uses the term "collection of documents", "Next"
uses the term "ordered sequence of documents", while "Preve" uses "ordered
series of documents". You should use only one term for the same thing. (BTW,
"Chapter", "Section", "Subsection" and "Appendix" use "collection of
documents", "Parent" uses "structured set of documents".)

[LinkTypes: Bookmark

"The title attribute may be used": "title" is marked up using a span element
with class='attributes', while other attributes' names are marked up as
"code" (e.g. "[...] the hreflang attribute [...]").


LinkTypes: required, prefetch, redirect

Just don't introduce, at least, "redirect".

6.1. Core Attribute Collection

"This attribute offers advisory information about the element for which it
is set."
Does this attribute offer information only about the element's contents or
also about other attributes used on the same element (e.g. edit, cite, href,

6.5. Hypertext Attribute Collection


"This attribute specifies the allowable content types of the relevant src
should be "[...] of the relevant [href] URI.", as "src" isn't declared in
this section.


"This attribute describes the relationship from the current document to the
URI referred to by the element."
There might be two URIs (in href and cite), so this should be something
similar to the description of "rev":
"[...] referred to by the href attribute of the element ."


"This attribute is used to describe a [reverse link] from the [anchor]
specified by the href attribute to the current document."
should maybe better (like rel):
"This attribute describes the relationship from the URI referred to by the
href attribute of the element to the current document."

accesskey, navindex

I would drop these attribute because I think that it's up to the user agent
to provide / create reasonable access keys and navigation orders through,
e.g., rel attributes.

6.5. Hypertext Attribute Collection, 6.6. Embedding Attribute Collection

href, src attributes
I think "href" and "src" are to presentional, maybe you could define them in
some other way, e.g.:

href: This attribute specifies a URI that offers advisory information about
the element for which it is set. (taken from "title")
 [This piece of information isn't required to understand the document, this
may also include, e.g., a "presentional" photo about "the atmosphere of a
party" which could also be embded (but replacing the content of the linking
element) "on load of the document". Or a webpage offering an index may also
be directly embded, ... This just doesn't reduce the possibilities of the
user agent to render the link.]

src: This attribute specifies the location of an external source that can
[replace] the contents of the element.

6.7. Image Map Attribute Collection

Drop this collection because it is completly presentional, authors may use
technologies as SVG or Flash to do something like image maps.

7. XHTML Structure Module

[footer PR #744: "[...] contain data that should be presented at the bottom
of content [...]"
We have address, nl, ..., so, what should this data semantical be?]

[security tag: "[...] elements that have security ramifications would be
rendered harmless [..]"
Such elements should alway be rendered harmless.]

7.3. The title element

Drop the title element and use h (maybe on html element) or meta data for

8. XHTML Text Module

[split text module
I would split text module in inline, block and maybe also "programming /
computer related elements" (like code, kbd, samp, var) which might not be
interesting for other environments (literature, newspapers, ...).

8.1. The abbr element

It's nice to have only one element for abbreviations and acronyms. Good

8.3. The blockquote element, 8.15. The quote element

When does inline quoted text become a block of quoted text?
This might also be done through <section><q></q></section> or (change model
of "q") <q><section></section><section></section></q>, no need for an extra
element for blocks.

8.7. The div element, 8.18. The span element

What's the difference between div and span, drop div (not span) because it
may be in a "conflict" with section (similar names),  span's name is a bit
more "presentional" and "universal" ("span over several words, sections,

8.12. The l element, 8.13. The p element, 8.17. The section element

When does a section become a paragraph, when does a paragraph become a
sub-paragraph (a line)?
You should only use one element (I propose "section").

When "l" means, e.g., a line of a poem, a verse of a poem, why hasn't a
strophe its own element?
A line is just a subunit of a poem, as a strophe is, why shouldn't we
represent them as sections.
Also, poem lines, poem verses aren't always presented as "lines", they may
also be separated by a slash (line 1 / line 2 / line 3 / ...) in a "inline
context", ...

When you use a "line" of computer code, it isn't the line that matters, it's
the character (LF and/or CR) that ends a line (in a "data stream").
So you should "present" computer code using pre or xml:space.

And of course, you can also use lists for (some types of) "lines".

8.14. The pre element

Can't we do the same with xml:space on any element?

8.9. The heading elements

Drop h1-h6, they are no more needed.

8.10. The hr element

Drop hr, it's only presentional

8.19. The strong element

Remove strong, or they must be clear rules when to use strong instead of em
(like it was('nt) with acronym and abbr).

8.20. The sub element, 8.21. The sup element

Remove them, they are presentional. For charcaters (scripts) that need to be
"sub" or "sup", there should be special characters in Unicode (I know this
isn't Unicode, but: ^2 = , ...). With these special characters (if they
have a clear meaning, e.g. POWER 2), it should be simpler for user agent to
replace the with, e.g., ^2.

9. XHTML Hypertext Module

Drop this modules, hypertext can be used on every element.

10.2. The nl element

"On visual user agents, the default presentation behavior is as follows:

This is too presentional, it may be there as an example only.

10.5. The label element

Replace it with using h.

11. XHTML Client-Side Image Map Module

Drop it, see above (attribute collection).

14. XHTML Object Module

Drop it, src can be used on any element to provide "object alternatives to
the elment's content".

15. XHTML Scripting Module

This should also be done through "link", "src" or "href"

16. XHTML Style Sheet Module

Just don't include internal style sheets.

17. XHTML Tables Module

17.2. The col and colgroup elements

There should something similar to thead, tbody, tfoot for colgroup, to group
heading information, ... also in cols.

[Maybe you could reuse the "span" element as "col" element, as col has the
same meaning for cols, that span has for text.]

Thank You for reading

 Rafael Gieschke
Received on Monday, 28 April 2003 14:31:40 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:17:44 GMT