Showing posts with label HTML5. Show all posts
Showing posts with label HTML5. Show all posts

Monday, July 25, 2011

HTML5 tags, metadata and microdata

I moved the JavaScript-enabled XHTML+RDFa home page for Aule Browsers to HTML5 today.  About half of the Dublin Core metadata was lost in doing so, as was Common Tag for Freebase.  All that was gained was the addition of Schema.org microdata.

Whatwg.org appears to have little sympathy for RDF where it is used.  There is a tendency to mistake RDF for one implementation in XML. Is that the source of this conflict? I see very little sign that Schema.org and microdata are a step towards the semantic web - and HTML5 itself offers as semantic markup such gems as "section" and "article".

The present deficiencies of the types offered by Schema.org seem to reflect geek mentality. poem is missing.  Language sprawls across natural language and programming language.  Just find a peer-reviewed linguistics article that argues for Python as a language and which article is then referenced with approval by a linguist ranked in the top twenty in that branch of science in a published book on his or her own specialty.  Oh, yes, "semantics".

Just try to start a dinner conversation on a blind date with a non-geek with
def nonsense()
WHITESPACE WHITESPACE pass
Pass the salt?

Where is that Pimsleur CD for Perl, lingua franca of the web ...
Sorry, but was that the punch line, or did you leave off a parenthesis?
XML-me, I missed a bracket when I removed lint from my one sport jacket.
That w3.org is not in the grip of Microsoft is no great consolation if it now guided by Google.  I mean,
(Google + (Bing + Yahoo!))  or, in summation, search for content trumps content.

My vote is for rich content; let the search engines be the ones who play catch-up.

Monday, April 5, 2010

David Frum on the iPad

I posted this note over at his "forum".
The eBook mobile device has a few challenges: one is the rejection of commercial formats by the free-software movement. The MIT web formatting language CURL (www.curl.com) was spun-off as a company and so got no traction: but it offers a great deal more than PDF. It is more than just text and imaging: we need options to comapre versions/editions/translations and options to interpolate footnotes or side-pane footnotes and to link to eTexts from footnotes and options for our marginalia, notes and annotations. Microsoft OneNote 2007 on a netbook is about the best thing at the moment.


A thinking person's eBook viewer is NOT a web browser. I ahve some notes on the topic over at aule-browser.com

Other formats include Djvu, Daisy, JBig2, MobiPocket, Plucker

EPub is basically just a ZIP file

The Curl format is a much richer alternative to HTML5 or the Literature Markup Language due to the availability of macros and user-defined formats.

Another long-shot: www.rebol.com or www.rebol.net with the arrival of Rebol3

The most neglected way to index and explore text: ICON as in arizona.edu/ICON or ObjectIcon at code.google.com (from the folks who brought SNOBOL to the humanities.)

Friday, October 30, 2009

HTML 5, Curl and presentation

I thought it might be helpful to post the "tags" used as text-formats in the Curl language (at least those in the package CURL.GUI.TEXT-FORMATS)

Curl also allows the creation of user-defined formats, text-procedures and macros so the power of Curl for markup and markdown goes far beyond these expressions.

HTML 5 is expected to drop so-called "presentation" elements such as those using the tag "br". CSS and possibly XSLT are required to achieve such simple text formatting for "presentation".

In Curl, text itself is treated as a Visual object and with Curl 7.0 additional classes and expressions were added to strengthen this alternative one-language approach to presenting structured content whether in the browser or in a desktop application.

The 6.0 version of the text-formats can be found on-line at developers.curl.com (click "show" in the upper-left corner of that page to open the full browser view.)

The Curl 7.0 list of text-format tags follows:
big
blockquote
bold
br
browse-url
bullet
cell
center
copyright
default-browse-url
definition
definition-list
degrees
destination
elaborate
em-dash
en-dash
enumerate
full-copyright
full-rte-copyright
header-cell
header-row
heading
hrule
huge
image
italic
item
itemize
left-justify
line-through
link
marker
monospace
no-break
numbered-heading
page-break
pre
quote
registered-trademark
right-justify
row
small
subscript
superscript
table
term
text-part
tiny
title
trademark
underline
In the above list "link" corresponds to the HTML "a" anchor tag.

{text-part some text of ours} is particularly interesting as an example of the difference between the Curl approach and the X/HTML approach to web content.  When the expression is evaluated by the Curl runtime engine, a Visual object is returned which can then be passed into other expressions, macros or procedures.
In those contexts, the .add method - common to all descendents of Visual - can be used as a common protocol without concern whether the original string content was a String, a StringBuf or other type of text object, as in
set sample-visual = {text-part some string content of yours}
{some-action sample-visual}
|| comment: where in {some-action } we would find
   {visual-parameter.add "more content"}
Curl 6.0 added a library for separation of style-sheets and Curl 7.0 incorporated this into the default packages so that each of the above elements can be styled just as one would use CSS to style HTML.

In addition to text-formats, Curl includes text-procedures such as {text } and {paragraph } and a wide range of predefined paragraph-formats.  Further, the standard libraries include {TextArea }, { TextFlowBox } which go beyond markup Visual objects to being full-blown GUI widgets (Graphic objects) such as usually would require JavaScript or the like and which respond to UI events such as mouse clicks.