Held Thursday, September 2, 1999
- Logical vs. physical markup, revisited
- Can the appearance of documents tell us about their intent?
- Can/should the ``same'' document have multiple appearances?
- Can/should the ``same'' document have multiple contents?
- Designing a language for marking styles
- Cascading style sheets, a Web standard
- Reading Assignment (due Tuesday, September 7):
Copyright in the
- Questions due by noon, Monday, September 6
- Reading Assignment (due Tuesday, September 7):
Advice for Students
on Citation and Scholarship
- Don't forget to get your essays in today.
- We've extended the due dates for
- Don't forget that you have appointments with Sam today and tomorrow.
- Sean and Ellen need to reschedule.
- Note that you will have classes on Monday: Grinnell does not
- Jamal Rogers, one of my upper-level advisees, may be performing
in Bob's on Sunday night at 8:00 or 9:00 p.m.
- In our previous discussion, we identified a number of arguments
for logical formatting and for physical formatting.
- Authors might prefer physical formatting because they see a need
for close control over the document's appearance.
- A clever author might use layout, font, and color to convey
- Readers might prefer logical formatting because their choices and
assumptions might clash with those of authors.
- A central idea in hypertext is ``reader empowerment''
- Some readers prefer sans-serif fonts, other prefer serif fonts.
Authors typically don't care how readers read.
- Some readers need larger fonts.
- Some readers are color deficient (e.g., SamR) or color blind.
- Authors might prefer logical formatting because it can support
more careful writing (``what role does this stuff play'') and
because it makes obviates the need for authors to identify and
use particular stylistic guidelines.
- Once you've said something is a book title, it's up to someone
else to decide what book titles look like.
- Readers and authors might prefer logical formatting because it
helps handle the typical strategy of ``one appearance can mean
multiple things''. Consider
In ``Beowulf gives Grendel a surprise'',
Beowulf is a noun
functioning as the subject.
(Paraphrased from Dobbs, Dr. Syntax, I: Syntax: Models)
- Authors might prefer logical formatting because it helps them
write ``publication-independent'' papers. That is, when I
write a journal article, I eventually need to reformat it to
the specifications of the journal. Journals differ on
- Margins and type size of the abstract
- Justification, type, type size, and capitalization of section
- Separation and indent of paragraphs
- And much much more
- If I write in a logical format and the journal provdes an appropriate
translation, everyone's happy.
- Unfortunately, early versions of HTML provided only limited support
for either kind of formatting.
- The logical tags were limited to:
P : paragraph
EM : emphasized
STRONG : strongly emphasized
OL : ordered list
UL : unordered list
LI : list item
CODE : program code
DL : description list (term list)
DT : description term
DD : description definition
A : anchor (link or destination)
ADDRESS : address of author
META : information about the document
TITLE : title of the document
CAPTION : the caption of a table
BLOCKQUOTE : quotation
HEAD : information about the document
BODY : the contents of the document
- The physical tags were limited to:
B : bold
I : italic
U : underline
TT : typewriter face
BR : line break
PRE : preformatted text
- Some seem to be either/both
HR : horizontal rule
IMG : an image
- As HTML evolved, tags for describing tables were added.
- These can play logical roles.
- They are also used for physical formatting. For example, you can
use tables to get things side-by-side.
- Frames were also added, but they are an abomination.
- Surprisingly, although HTML initially had a simple logical emphasis,
it quickly gained physical tags.
- It also gained parameters to the logical tags that had only physical
effect. For example
<BODY bgcolor="black" text="red">
- At the same time, browsers provided only limited support for readers
to customzie their environments.
- Readers could choose a default font and size, but that's about it.
- The HTML community began to consider ways to extend HTML to
better support authors (who wanted or needed tighter control
over appearance or support for logical authoring) and readers
(who also wanted additional control).
- We'll spend a few minutes considering how we might change
HTML to better support readers and authors.
- What are the requirements?
- How might we meet those requirements?
- Eventually, the community settled on a notion known as
cascading style sheets (CSS).
- Authors and readers can carefully define the appearance of
different document components.
- Authors can create variants of existing components.
- Using the cascading style sheet model, authors can define the
``type'' of a particular element. For example,
<P class="abstract"> : for a paragraph that
serves as an abstract
<em class="bluish"> : for emphasized text
that might appear in blue.
- Authors get to choose the classes; there are no (or almost no)
- Authors then define style sheets that indicate how
the portions are displayed.
- Each entry in the style sheet has the following form:
- For example,
- There are some more sophisticated things you can do with style
sheets, particularly contextual styles. However, they are
beyond the scope of this class.
- Once you've created a style sheet (saved as
you add something like the following to the head of the document
to use it:
<LINK REL="stylesheet" TYPE="text/css"
- What kinds of things can you affect? Nearly everything.
- The font, color, background and such for your text.
- The layout of each piece of text (relative to the previous text or
to the page as a whole).
- There are other ways to incorporate styles in documents. We'll leave
those to your own investigations.
You will have the opportunity to explore cascading style sheets in
Monday, 22 March 1999
- Created as a blank outline. (Can you tell
what I was doing during Spring break?)
Wednesday, 1 September 1999