Re: [css-gcpm] mostly editorial comments on footnotes

On 1/29/14 4:53 PM, "Chris Lilley" <chris@w3.org> wrote:

>Some comments on CSS Generated Content for Paged Media Module
>Editoršs Draft, 28 January 2014
>
>1) content-list definition
>
>http://dev.w3.org/csswg/css-gcpm/#string-set
>
>"content-list = [ <string> | <counter> | <content()> | attr() ]+"
>
>"<counter>
>    A counter() or counters() function, as described in [CSS21]."
>
>
>For consistency, maybe indicate the functional nature more clearly
>with
>
>"content-list = [ <string> | <counter()> | <counters()> | <content()> |
>attr() ]+"
>
>(no strong preference, just makes it clearer in the content-list
>definition that these are functions).

Fixed.


>2) entry value and exit value
>
>These terms are frequently used. They should be defined clearly in one
>place which can be linked to (or, if defined in another spec, linked
>to that definition and that spec added to the normative references).

These are now marked as definitions. I'm also working on another example
to clarify these terms when elements fragment across page boundaries.


>In particular, for all properties that may also be used in non-paged
>media, the definitions of entry value (at least) should be clearly
>defined in non-paged contexts.

I'll have to think about this. I can imagine something like a running head
in a scrollable context (via position: fixed) where the content depends on
what document section is currently visible.


>3) Requirements for Latin Text Layout and Pagination
>
>This specification is referenced in passing (as a definition od
>"complex object") but that specification is not in the list of
>references. It should be added as a normative or informative reference
>depending on how it is used here and subsequently.

Latinreq will be published as a first public working draft in W3C space on
March 13. I've updated this section so it will appear in references, and
link to the FPWD.


>4) footnote-only pages
>
>  "unless the page contains only footnotes (as may happen on the
>  last page of a document). "
>  http://dev.w3.org/csswg/css-gcpm/#footnote-area-size
>
>this seems to conflict with
>
>  "A footnote is created when such content moves to the bottom of the
>  page, leaving a reference indicator."
>  http://dev.w3.org/csswg/css-gcpm/#footnotes
>
>because there would be no content to generate the footnotes. Or at
>minimum, there would be the footnote call outside the footnote area.
>
>Unless this is intended to cope with page-overflow of the footnote
>area. Which could of course happen, and seems to be allowed by
>footnote-policy:
>http://dev.w3.org/csswg/css-gcpm/#footnote-policy
>but would not be restricted to the last page of a document.

Sometimes there's not enough room on a page for all the footnotes that are
referenced on a page. The page size is finite, and the footnote area
itself can be constrained by the max-height property. So the footnote area
can overflow, and footnotes (or parts of footnotes) end up on page(s)
after their callout.

In really difficult circumstances, footnote content may accumulate as the
end of the document is reached. In this case, the least worst option is to
place all the remaining footnotes at the end of the document, and so the
result may be a page (or pages) containing only footnotes. And if all the
other document content has already been rendered, there's no reason to
restrict the height of the footnote area, and it's OK to fill pages with
leftover footnotes.


>5) Footnote counter
>
>  "3. The footnote counter is incremented. "
>  http://dev.w3.org/csswg/css-gcpm/#creating-footnotes
>
>I was not at first not seeing text that defines how the footnote counter
>is
>initialized, and also whether it can be reset (for example for
>per-chapter footnote numbering).
>
>Suggest hyperlinking "footnote counter" in step 3 to
>http://dev.w3.org/csswg/css-gcpm/#footnote-counters

Done.


>6) name of block footnotes
>
>  "footnote
>    each footnote element is placed in the footnote area of the page
>    as a block element.
>
>  inline-footnote
>    each footnote element is placed in the footnote area of the page
>    as an inline element."
>    http://dev.w3.org/csswg/css-gcpm/#footnote-types
>
>Would it be correct to conclude that most footnotes are expected to be
>block and thus, footnote (for brevity) rather than block-footnote?

Yes.

>
>Potential drawback, harder to add new types (column footnotes, etc).
>block-footnote might be a better term, albeit longer.

True. There has been some subsequent discussion on IRC about having only
float: footnote, and making the block/inline distinction via the footnote
area. I'll be working on this.

>
>7) footnote-call default styling
>
>  "By default, the content of this pseudo-element is the value of the
>  footnote counter, styled as a superscripted number. "
>  http://dev.w3.org/csswg/css-gcpm/#footnote-call
>
>Could I infer then that example 9 is an addition to the default user
>agent stylesheet?
>
>  ::footnote-call {
>      content: counter(footnote);
>      vertical-align: super;
>      font-size: 65%;
>   }
>
>A problem here is that this seems to formalise faux-superscripts. This
>is unfortunate, at a time we are finally dealing with faux-bold,
>faux-italic and faux-smallcaps; Unicode has superscripted number
>characters, and opentype features can be used to produce proper
>superscripted numbers and other types of superscripted symbols.

Excellent point. I've updated the example to use font-variant-position:
super. I did note that CSS Fonts Level 3 recommends against using this
directly in user agent stylesheets [1]. I'll explictly add a UA stylesheet
to the spec that uses the specified fallback.

>8) footnote-policy: auto wording
>
>  "A footnote body may never be placed on a page before the footnote
>  reference."
>
>That sounds as if it should be normative, and testable. Thus, please
>reword as
>
>  A footnote body must never be placed on a page before the footnote
>  reference.

Fixed.

Thanks so much for the comments!

Dave


[1] http://dev.w3.org/csswg/css-fonts/#propdef-font-variant-position

This may contain confidential material. If you are not an intended recipient, please notify the sender, delete immediately, and understand that no disclosure or reliance on the information herein is permitted. Hachette Book Group may monitor email to and from our network.

Received on Tuesday, 4 March 2014 20:12:20 UTC