[uf-discuss] Citation Straw Proposal II (Recap)
brian.suda at gmail.com
Thu May 4 19:28:34 PDT 2006
This is a "quick" recap, point-by-point of the issues brought-up with
the Straw proposal II.
"I particularly like the use of hCard in this context (although I
think it's critical that we use more granular n attributes -- there
are just too many ways to mark up a citation)."
--- I have edited the Citation Example to explicitly use the
class="n". When ever a microformat is used within another, the full
spectrum of properties is possible.
"IMG -- Off the top of my head, I don't see much need for this... What
was the inspiration for this?"
--- I think there is a very fine line between something like Dublin
Core descriptive metadata and Citation metadata. DC has an IMAGE
property, so does hReview (along with hCard as PHOTO and LOGO). It
probably is not useful for a citation because the point of a citation
is to get you back to the original source, including IMAGE would not
add much, but i floated the idea because some one else might disagree.
There was a Mention of why the <cite> element was not used in the examples.
--- That is probably for two reasons, i don't want to get hung-up one
people thinking it HAS to go on <cite> or any other specific element
(people get confused and think that microformats HAVE to be on spans
and divs, which is not true at all!) The other reasons are in the W3C
examples and explanation of <cite>.
CITE: Contains a citation or a reference to other sources.
That needs further discussion on when and where to use that element.
One of the objectives of the citation microformat is to be able to use
citations as a bibliography AND inline a blog post. So in english
prose you might be putting more or less text than needed if you force
the root-element to be <cite class="cite-microformat">This is a great
book [...], you should read it</cite>. Having that all wrapped in
<cite> doesn't seem right.
Class="format" was used incorrectly in the straw example. I put <span
class="format">book</span> the format is more of the physical medium,
so it should be more like "Hardback", "Softback", "Audio", "Video"
journal implied schema:
--- i guess i must have just missed this one. i added this to the wiki.
<span class="title"><abbr title="book" class="format">Foo Bar</abbr></span>
--- The reason this might seem a bit strange is because we are
cramming a lot of information on the term "Foo Bar". This is the
Title, so it is wrapped in the <span class="title">, but it is also in
the format of a book (which is incorrect, see the above point). But we
don't want the word "Book" to be part of the title, so we nest an
<abbr> element around the text "Foo Bar". Then when the transforming
app comes around it will find the title:Foo Bar and the format:book.
(i hope this makes a bit more sense? The idea is in references to the
way TYPES are handled in hCard)
Author/Editor/Translator should be a priority.
--- This has been recently explored as possibly using the hcard ROLE
property. This is a very interesting idea. Basically, since we are
requiring that "creators" be marked-up as hCards we can use those
properties in both realms (vCard and Citation). Here is a simple
<span class="fn">Brian Suda</span>
<span class="fn org">XYZ Publishing House</span>
The question i am asking myself, is "Does role make sense outside of
the citation format?" "If i were to extract JUST the vCards is ROLE
still correct?" For the example above if i extracted the vCards it
would look something like:
FN:XYZ Publishing House
ORG:XYZ Publishing House
Outside of the context of the Citation, "Brian Suda" is still an
author (just think Venn Diagrams) Once you are an author, you can
never "Un-become" an author, so that statement is true. But does it
hold true for other property values, editor, translator, publisher,
copy-editor, etc.? (Before anyone argues the temporal aspect "i WAS a
translator, but don't do it any more" just remember that vCard has
things like TEL and ADR which are not infinatly true either) And then
should these be a defined list, or an opened-ended list and the
transforming application uses the ones it knows about and ignores the
rest (i think it probably should be a controlled enumerated list of
values gathered from the examples)
isPartOf seems pretty important
--- This one has got me thinking as well, and i'm interested in other
peoples' feedback. Since "Format" was incorrect, isPartOf is what i
intended. "This citation is part of a Book", etc. I am also wondering
if a sort of nested citation is a good idea or not? vCard has an AGENT
property, which is a vCard inside a vCard. We might be able to do the
same for a Citation.
A given article citation is part of a journal (which is just another
citation). The problem is that they would share ALOT of the same info
(PubDate, Publisher, etc) It would be difficult to publish an article
in a journal by two different publishers? (or i am off the mark here?)
So i'm not sure how much benefit there is in nesting citations.
the use of urls as IDs, and the use of additional protocols, such as
ISBN, urn:, etc.
Retrieved Date or Access Date
this was first brought-up in the straw example discussion, so the new
examples need to be folded into the schema. Can someone do a write-up
of WHAT IS ACCESS DATE and how it differs and why it is important.
Most of the cites between blogs, and places like Amazon.com, etc will
never really use ACCESS DATE.
If i am missing something let me know. Part of the push for this is
that the last two weeks of May I will be co-presenting several
microformats panels (shameless plug) at XTECH, www2006 and Refresh
Edinburgh. So this will be a great opportunity to show-off the straw
citation microformat to alot of other smart people and get their
feedback and to make them aware of our efforts.
Thanks for everyone's help and feedback.
 - http://www.w3.org/TR/html4/struct/text.html#h-9.2.1
 - http://microformats.org/wiki/hcard#type_subproperty_values
More information about the microformats-discuss