2 Dec 18:26 2009

### Re: Announcement: updated XHTML Strict writer

> On 2009-11-30, Alan G Isaac wrote:
>> <div class='footnote'><span class='label'>
>> and
>> <div class='citation'><span class='label'>
>> and let me style them.

On 11/30/2009 2:42 AM, Guenter Milde wrote:
> I'd rather use a styled list, so that it is still recognizable without CSS.
> Suggestions and patches welcome.

1. You mean to use a list for the citations, right,
not for the footnotes?  For footnotes, a list structure
does not make sense.
2. Even for citations, I think this is the wrong structure.
As for being "recognizable" without CSS, I think citations
will be adequately recognizable if each citation is put in a DIV
element: the presence of the label will provide a clear

I assume the list structure you might have in mind for
citations is the definition list (since the markers
will be annoying and confusing for an ordered or unordered
list).  Right?  This has some plausibility.
However the usual thing I wish to change in
citation display is to hide the labels altogether.

Alan Isaac

------------------------------------------------------------------------------


3 Dec 03:39 2009

### Re: Announcement: updated XHTML Strict writer

On Wed, Dec 02, 2009 at 12:26:29PM -0500, Alan G Isaac wrote:
> > On 2009-11-30, Alan G Isaac wrote:
> >> <div class='footnote'><span class='label'>
> >> and
> >> <div class='citation'><span class='label'>
> >> and let me style them.
>
>
>
> On 11/30/2009 2:42 AM, Guenter Milde wrote:
> > I'd rather use a styled list, so that it is still recognizable without CSS.
> > Suggestions and patches welcome.

Shouldn't citations be structured text as in
http://www.w3.org/TR/html401/struct/text.html#h-9.2 or at least go into blockquotes?

Cheers,
Harald

------------------------------------------------------------------------------
a free event focused on virtualization and cloud computing.
http://p.sf.net/sfu/redhat-sfdev2dev

3 Dec 04:23 2009

### Re: Announcement: updated XHTML Strict writer

On 12/2/2009 9:39 PM, hollow wrote:
> Shouldn't citations be structured text as in
> http://www.w3.org/TR/html401/struct/text.html#h-9.2  or at least go into blockquotes?

I believe the CITE element corresponds to an
reST citation reference, not to an reST citation.
One might argue for the label to be in a CITE element. Maybe.

A BLOCKQUOTE elementwould be completely inappropriate, imo.

But a DIV element with an appropriate class works great.

Alan Isaac

------------------------------------------------------------------------------
a free event focused on virtualization and cloud computing.
http://p.sf.net/sfu/redhat-sfdev2dev

3 Dec 10:06 2009

### Re: Announcement: updated XHTML Strict writer

On 2009-12-02, Alan G Isaac wrote:
> On 11/30/2009 2:42 AM, Guenter Milde wrote:

>> I'd rather use a styled list, so that it is still recognizable without
>> CSS. Suggestions and patches welcome.

> 1. You mean to use a list for the citations, right,
>     not for the footnotes?  For footnotes, a list structure
>     does not make sense.

Why not? Doesn't both, the reST markup and the HTML or PDF output
of the footnote text look very much like a list?

But the more important point is that semantically, both citation entry
and footnote text represent a

term (label) <-> description

relationship which in HTML is represented by a definition list.

> 2. Even for citations, I think this is the wrong structure.

I always think of a "References" or "Bibliography" section as a *list* of
references not as a generic text block of references.

>     As for being "recognizable" without CSS, I think citations
>     will be adequately recognizable if each citation is put in a DIV
>     element: the presence of the label will provide a clear



3 Dec 15:54 2009

### Re: Announcement: updated XHTML Strict writer

>> On 11/30/2009 2:42 AM, Guenter Milde wrote:
>>> I'd rather use a styled list, so that it is still recognizable without
>>> CSS. Suggestions and patches welcome.

> On 2009-12-02, Alan G Isaac wrote:
>> 1. You mean to use a list for the citations, right,
>>      not for the footnotes?  For footnotes, a list structure
>>      does not make sense.

On 12/3/2009 4:06 AM, Guenter Milde wrote:
> Why not? Doesn't both, the reST markup and the HTML or PDF output
> of the footnote text look very much like a list?
>
> But the more important point is that semantically, both citation entry
> and footnote text represent a
>    term (label)<->  description
> relationship which in HTML is represented by a definition list.

Actually the spec is clear that no one-to-one relationship
is modeled by definition lists: a term may have more than
one definition, and more than one term may share a definition.
So this justification is somewhat spurious.  If we really
believed that, reST would not have a separate syntax for
citation references and citations, since it handles definition
lists (with links to the label) just fine.

Furthermore, a footnote clearly is not a list,
and so creating a list container to hold is
is semantically wrong.



4 Dec 00:28 2009

### Re: Announcement: updated XHTML Strict writer

On 2009-12-03, Alan G Isaac wrote:
>>> On 11/30/2009 2:42 AM, Guenter Milde wrote:
>> On 2009-12-02, Alan G Isaac wrote:
> On 12/3/2009 4:06 AM, Guenter Milde wrote:
>>> 1. You mean to use a list for the citations, right,
>>>      not for the footnotes?  For footnotes, a list structure
>>>      does not make sense.

>> ... semantically, both citation entry
>> and footnote text represent a
>>    term (label)<->  description
>> relationship which in HTML is represented by a definition list.

> Actually the spec is clear that no one-to-one relationship
> is modeled by definition lists: a term may have more than
> one definition, and more than one term may share a definition.
> So this justification is somewhat spurious.

But the one-to-one relationship is possible as well. And a definition
list is clearly the best suited HTML element to represent such a
relationship. In any case closer than span and div.

> Furthermore, a footnote clearly is not a list, and so creating a list
> container to hold is is semantically wrong.

Footnotes are *listed* at the end of a page or document.

> ... However in contrast
> with LaTeX, HTML does not allow us to simply define our
> own list-like environments. Instead only three predefined


4 Dec 03:09 2009

### Re: Announcement: updated XHTML Strict writer

> On 2009-12-03, Alan G Isaac wrote:
>> Furthermore, a footnote clearly is not a list, and so creating a list
>> container to hold is is semantically wrong.

On 12/3/2009 6:28 PM, Guenter Milde wrote:
> Footnotes are *listed* at the end of a page or document.

That is true in some contexts and not in others.
In the context of reST, I find that an odd claim.
If reST implements an endnotes directive, you would
have a stronger case to handle endnotes as a
definition list.  (Looking forward!)

> On 2009-12-03, Alan G Isaac wrote:
>> Note that LaTeX has enumerated
>> lists, itemized lists, and definition lists, yet it does
>> *not* use any of these for citations.

On 12/3/2009 6:28 PM, Guenter Milde wrote:
> Still, "thebibliography" is a list, based on the generic "list"
> environment.

Yes, as I said, "in contrast with LaTeX, HTML does not allow us to
simply define our own list-like environments."

On 12/3/2009 6:28 PM, Guenter Milde wrote:
> and not try to style a paragraph to get a list of citations or footnotes.



4 Dec 10:43 2009

### Re: Announcement: updated XHTML Strict writer

On 2009-12-04, Alan G Isaac wrote:
>> On 2009-12-03, Alan G Isaac wrote:
>>> Furthermore, a footnote clearly is not a list, and so creating a list
>>> container to hold is is semantically wrong.

> On 12/3/2009 6:28 PM, Guenter Milde wrote:
>> Footnotes are *listed* at the end of a page or document.

> That is true in some contexts and not in others.

Footnotes are always listed. The layout of this listing may differ (there
is, e.g., a LaTeX package to collect footnotes in one paragraph) and
sometimes the listing holds just one item, but there is always a label
followed by content.

Wikipedia uses an ordered list for footnotes,
"It is the perfect choice also for a footnote section."

> In the context of reST, I find that an odd claim. If reST implements an
> endnotes directive, you would have a stronger case to handle endnotes
> as a definition list.  (Looking forward!)

Odd? Current reST syntax requires you to explicitly list footnotes at
the place where they appear in HTML. But why should the output be
different for automatically collected footnotes compared to
manually placed ones?

> On 12/3/2009 6:28 PM, Guenter Milde wrote:
>> Still, "thebibliography" is a list, based on the generic "list"


4 Dec 16:12 2009

### Re: citations in XHTML Strict writer

>> Alan:
>>>  To repeat, having a label is **not** fundamental to the
>>>  citation concept.  Overlooking that misleads you, imo.

On 12/4/2009 4:43 AM, Guenter Milde wrote:
> The citation (bibentry) element in the Docutils doctree has an obligatory
> label.

This is really irrelevant, right?  I am talking about the *citation
concept*, not the implementation details.  As evidence I can point
to almost any journal in my field.  Naturally different fields follow
different practices, but that just makes my point.

On 12/4/2009 4:43 AM, Guenter Milde wrote:
> Suppressing a list label with CSS is simple.

Of course.  If you use defintion lists for citations,
I will plan to do something like::

dl.citation dt {display: None}

So while the definition list structure is conceptually wrong, it *may*
prove to be an fairly innocuous choice.

Alan

------------------------------------------------------------------------------
a free event focused on virtualization and cloud computing.


4 Dec 16:17 2009

### Re: bib4txt

http://code.google.com/p/bibstuff/source/browse/#svn/trunk/examples

hth,
Alan

------------------------------------------------------------------------------