Bibliography lists

11 posts / 0 new
Last post

Three questions.
The accessibility guide mentions using un-ordered lists or definition lists. Is there a reason for this? Bibliographies are usually alphabetically sorted and we use ordered lists in our production, optionally broken into alphabetical sections and reference types (book, internet, etc) for large and complex reference lists.
I can't see how an academic reference list can use a definition list unless the alphabetic sort term is used for the term. Can we assume this is a non-starter with real world content?
We use NLM inspired detailed semantic span tagging in reference items as metadata processing targets. A reference item can contain up to 12-14 span statements. Does this density of span elements have a derogatory effect on accessibility?

Sorry about the double entry. I had a fight with my browser!

Re Q1: It mentions unordered lists because alphabetized presentation of the information isn't necessarily the same thing as ordered presentation. A similar question was just raised in the index work that's going on, namely that there are other ways of presenting the information in an index. A-Z is typical, but you could equally validly present the index Z-A without losing information, or by all proper names first, etc. The same applies to bibliographies. Assuming they eventually get the same treatment, you could re-order by the referenced work, the publisher, etc. without any change in meaning. The first entry in a bibliography is not numerically important, especially if you subsequently turn numbering off for presentation. They're not like a table of contents where if you re-order the information you do lose meaning (except perhaps in the odd cookbook).

Re Q2: I've seen all kinds of oddness in bibliographies, but if a definition list doesn't work for you, I can't argue with not using them. I have seen bibliographies ordered with author name followed by a listing of all works (not one entry per). I've also seen web pages cited with offset text summaries following explaining why, sort of like a quasi-annotated bibliography within a bibliography. These layouts lend themselves to dl markup, and lists allow easy traversal by someone using an assistive technology.

Re Q3: No, the number of span elements shouldn't affect the accessibility. I've not heard of any negative effects on text-to-speech playback, and a word-level synchronized media overlay would have spans on every word in the content, for example (although it will be interesting to see if epubcfis can do away with that headache).

Thanks for the comprehensive reply, although the concept of an un-ordered index leaves me flummoxed. A-Z or Z-A it is still ordered. I always interpreted OL as being a defined sequence which must be maintained, and UL as it doesn't matter what the order is. The reason this distinction is important is for post-processing content, remixing content together, etc. However I understand that is not an issue for the IDPF.

Just one last question on this. Will assisting technologies do something wrong or fail if there is an OL instead of a UL? (We will process these structures to the UL requirement at format generation time). This is more a general question on the strictness of behaviour.

No, the primary need for accessibility is to use some kind of list structure, as that's what facilitates navigation. If you choose to use an ordered list, it's not going to have an appreciable impact on the accessibility of the publication.

A-Z or Z-A are technically just sortings in these cases, and there are other ways you could sort the data without loss of information. That's why I wouldn't see the use as incongruent to what you note, if perhaps we don't agree on the mutability of the data. The original alphabetic sorting may be preferred, but the information could be re-sorted at any point in the chain without fundamentally altering how a reader would understand the information in those sections.

Or perhaps another way of putting it is that you never have to qualify the order of information in an index or bibliography. The function of the section is understood regardless of the presentation of the data. The same is not true of real ordered lists, where even something as simple as a reverse sort could completely change the meaning and would likely require rewriting the surrounding prose for the new order to make sense to the reader.

But as I said at the start, the goal for accessibility is simply to have a list structure. If a good case gets made later why ordered lists should be preferred instead, I'm open to change, but at this point the use of unordered matches the work going on elsewhere in the IDPF.

Hi All,

Sub: ePub Footer display in each page:

Please help me to get the solution for the below query:
I have used footer tag in html and defined the value in CSS. when i load the epub file in Ipad, i can see the footers appears only in first two pages of each chapter and the remaining pages seems to be disappears in ipad. Could you please guide me to fix this issue.

Thanks in advance...


How are you affixing the content, using position: fixed?

If so, there's no guarantee what affect that will have in any given reading system, and it's use is discouraged in the specifications for this reason (see the first bullet in section 3.3.1).

In a paginated, reflowable context, the header and footer elements typically represent the very first and very last content in the document, nothing more (except perhaps for the semantics of being header and footer material to the main text). The elements are not designed to render on each page of content as the reader moves through a book.

If you're looking at running headers and footers, have a look at the oeb-page-head and oeb-page-foot properties in section 3.3.11. Unfortunately, they're not well supported in the wild at this time.

I believe the advanced adaptive layout work could help define a page template with header/footer regions for content, but I don't know of any support for those layouts at this time.

Thank you for your comments. So i assusme its not possible to make the footer in each page.

But my query is, how its coming correctly in two pages and why not the remaining pages?

Kindly let me know if you have any ideas!

Thank you,

Some quirk of Apple's rendering engine, no doubt.

Since two-page spreads are not uncommon for viewing, those two initial pages may be rendered in similar fashion, with the fixed content, but all additional pages without. Only Apple would know.

As I understand the issue, it's not a trivial task to render and paginate the content and at the same time discover what should be affixed and where and also try to re-incorporate that content into the rendering of every paginated view. Fixed content really only works well in scrolled environments.

There's no universal process for rendering content across reading systems, but if you want a general idea of how it's done, have a look at the refactoring page on the Readium site.

Ok thankyou so much for your comments. Just keep update me if you find anything regarding this...



Please find the below points and suggest? Shall i get any other options?

1. Footer(image format) in ePub2 is possible?
2. We have used image logo for footer.
3. I can see the footer in two pages of each chapter but the content seems to be hiding in each pages where footer appears. Is there any option to fix the bottom margin? Also suggest how to separate with space in between "the body content" and "footer" per page.
4. We have used div id="container" for body content and div id="footer" for footer in each html file and defining the rules in CSS.


Secondary menu