Vol. XXIV, No. 3January, 2012

 

Articles in Vol. XXIV, No. 3

Website Review: Ara Pacis Augustae
In-depth visual documentation.
-- Martin Beckmann

Websites as Stable Resources
Too many have disappeared.
-- Harrison Eiteljorg, II
§ Readers' comments (as of 2/3/2010)

Website Review: Dikili Tash
A good beginning, but only a beginning.
-- Andrea Vianello

Evolving Web Standards: a Blessing and a Curse
Why eliminate good, useful tags?
-- Harrison Eiteljorg, II
§ Readers' comments (as of 2/3/2010)

Project Publication on the Web - Addendum
Making it clear that a final report is just that.
-- Harrison Eiteljorg, II


To comment on an article, please email
the editor using editor as the user-
name, csanet.org as the domain-name,
and the standard user@domain format.


Index of Web site and CD reviews from the Newsletter.

Limited subject index for Newsletter articles.

Direct links for articles concerning:


CSA HOME PAGE

Search all newsletter articles.
(Using Google® advanced search
page with CSA Newsletter limit
already set.)


 

Reader Commentaries on and Responses to

Evolving Web Standards: a Blessing and a Curse
by Harrison Eiteljorg, II


This comment was made by Martin Greaney and posted on February 2, 2012.

I completely understand and sympathise with the view that constantly evolving standards can cause problems for the stability of long-term web content, but I want to suggest that the changes are more of a blessing than a curse, as is the separation of meaning (tags) from presentation (CSS).

The biggest advantage of this separation is that a browser (either in the sense of a human reader or a piece of software) finds it easier to deal with each part separately. This is done most critically in software used by those with visual impairments or complete blindness (screen readers are one example, though there are others). Visual style obviously makes very little sense in this context, but if the HTML correctly uses only tags denoting meaning, and not presentation, any browser can still get meaning from the page.

This helps not only screen-readers, but also e-readers and tools like Readability, or some yet-to-be-dreamed of technology. When your page is transferred to the other medium or device, only the meaningfully tagged HTML need go. The device can then deal with presentation in its own way, possibly influenced by user settings.

However, as someone who has used and developed for the web for most of a decade, I can see why ever-changing standards can be a problem. Something tagged 10 years ago may start to look odd in modern browsers, or not work at all. This is why standards are evolving with the meaning-presentation divide in mind. Over time, changes to the standards will be more backwards-compatible and less of a pain for writers.

So to summarise, it is in our interest as website owners to stick to meaning-only tags, and keep all presentation in CSS. It maximises accessibility (sometimes a legal requirement for government sites) by removing the reliance on visual appearance for meaning. It also helps interoperability, and it should help to future-proof us against the pain of evolving standards by separating appearance from HTML tags altogether.

 


A response from Harrison Eiteljorg, II, posted February 3, 2012.

Much as I would like to agree with the comments above, I have a basic difference of opinion.

I must first say that the issue of tagging for use by software for those with limited eyesight or any other purpose seems to me to be a red herring. There is simply no reason that software cannot deal with tags effectively, regardless of the way(s) those tags are included or referenced. So long as tags are well-defined, they can easily be ignored, attended to, or otherwise taken into account. Whether the tags are internal to the document or included by reference to a stylesheet is a matter of no consequence, and how they are interpreted is a matter easily determined by software. Indeed, current software transparently uses both HTML tags within a document and referenced stylesheets, and current software easily and transparently ignores undefined tags.

Now back to the basic disagreement. Is there really a hard-and-fast distinction between "meaning" and "presentation" in a document, or is that distinction essentially an artificial construct? I would assert that is is, in fact, a totally artificial distinction.

I have tried multiple iterations of what follows, but each time I look more deeply into the recommended coding, I find that the target has moved again; so I must resort to language that may seem terribly vague. Please bear with me. HTML is so fluid at this moment that it is simply not possible to make many declarative statements about it and expect them to be valid tomorrow. That simple fact is, of course, the core of the problem.

The two simplest examples of the unreal distinction between meaning and presentation were already noted in the article. The tags for emphasis (<em>) and strong (<strong>) are said to represent meaning, but they are, in fact, used to produce Italics or boldface. While one might make an argument that there really is meaning in a tag for emphasis or "strength" (stronger than simple emphasis), the only reason for two tags is to permit the author to use Italics or boldface and to specify which. Anther example is the tag that inserts a line feed but not an end-of-paragraph mark (<br>) into the text. It is, so far as I am aware, still in the pantheon, and its only purpose is to permit the user to force a line feed for the sake of appearance. (It is commonly used in CSA Newsletter articles when a caption requires multiple lines and would look unruly without forced line feeds.) Another example from the experience of the editor, is the use of font size and position (small, superscript) to indicate a footnote. This is, of course, an appearance and a meaning matter, but I am aware of no tag to indicate a footnote number, much less the footnote itself; so one must make do with the available techniques. Perhaps the most obvious silliness is the presence of multiple levels of headings (I believe there are six). Does anyone know what the difference is between and among them? Does anyone actually use more than one or two in a any given document? (The term anyone in the foregoing must be taken to mean anyone other than a professional website designer. There are people — this author among them — who create web documents regularly but only as a "sideline.")

I do not think further examples are necessary, but I would add that there are so many missing tags for meaning that the idea of tagging according to meaning seems ludicrous and perhaps a stand-in for those who have tired of seeking the Holy Grail. For instance, there is a single tag for citations (<cite>), but there is no distinction between/among book titles, article titles, songs, poems, magazines, newspapers, . . . . Similarly, I am aware of no tags for the author whose name may accompany a citation (though I have seen an example of an author's name being tagged as a citation). One could go on and on with a personal list of desired tags to make the system work (e.g., my desire for a <foreign> tag to indicate a foreign word, noted in the original article), but the point is not that some tags are good or others bad. Nor is the point that we need more or fewer tags. The basic point is much simpler: absent a vey good and compelling reason, we should not be producing "standards" that make existing documents obsolete. That is, however, what the HTML standards groups have done. They have, by removing support for specific tags, threatened any number of extant web resources without a compelling reason to do so. These standards bodies could simply require the continuation of support for old tags (while introducing new ones) and have no deleterious impact on extant documents. A cynic would argue that such an approach would remove the advantage of those who make their livings creating and updating websites. A realist must wonder if, in fact, these standards bodies do not need to keep changing things both to retain the advantages of the cognoscenti and to demonstrate their own importance. Imagine a group empowered regularly to update the alphabet used for American English so that spelling might be more predictable. Who would be able to write? And who would be able to read old things? Only in the area of technology are we willing supinely to cede such power to the proverbial powers that be.

One final note. How much time and money will it cost organizations like CSA to deal with their web documents so that they can retain their utility and then, secondarily, provide the sources for intellectual historians of the future as libraries do today? Is that really how we want to have our time, our efforts, and our funds used? I can speak only for myself and CSA when I say that I do not wish to to see our web documents become useless; nor do I wish to see time and money spent solely at the insistence of the gods of the internet, no matter how exalted or visionary they may be.

 

 

About this document:

Commentaries for the CSA Newsletter are assembled by the staff in cooperation with contributors. All are published with an assumption that there will be additions from time to time and are maintained, with the latest additions, at the CSA website. While additions are normal, changes other than corrections of typos or similar errors will rarely be made after publication. If any such change is made, it will be made so as to permit both the original text and the change to be determined.

Comments like those here are welcome, and comments, questions, concerns, and author responses will be published in these separate commentary pages regularly.