The web contains vast amounts of stuff said about the web. Its history, its philosophy, its politics. Perhaps even its contribution to questions such as "what is the answer to life, the universe, and everything?" Or at least to questions like "how can I communicate with my fellow human beings?"

This stuff includes lots of things that people have said about the web and its nature. Some of it is weird! Some of it simply needs interpretation. Here is my attempt.

"The web is not DTP"

If all you have is a hammer, everything looks like a nail. 


If all you have is a screwdriver, everything may get screwed up. 

Barry Pearson, 2004

It's only mark-up. It isn't sin! 

Barry Pearson, 2004

The aim of a web site is to communicate with the audience. Everything else is a means to that end.  

Barry Pearson, 2004

Designing with CSS is sometimes like barreling down a crumbling mountain road at 90 miles per hour secure in the knowledge that repairs are scheduled for next Tuesday. 

Al Sparber

Project Seven Development

Under the most rigorously controlled conditions of pressure, temperature, volume, humidity, and other variables, the organism will do as it damn well pleases. 

First Harvard Law of Behaviorism

Using tables for truly tabular data is actually quite rare online. In the immortal words of William Gibson, the street finds its own uses for things. The use the street found for tables is graphic design, not data structure. When you employ tables for layout, you can get away with minimal coding. Data tables require significant markup skills that take a long time to acquire. 

Joe Clark, 2002

"Building Accessible Websites", chapter 10, "Tables and Frames"

 Since 1994, HTML has been used a thousand times to control presentation for every one time it's been used to express meaning alone. The issue of separate presentation and content in HTML itself is long closed. It was a noble idea but it's not what happened....

While trying to make visual the intricacies of HTML in a form useful to web designers I finally realized HTML is all presentation. For years I'd preached its intent was to code structure and let automation decide presentation. But the data from a spider program I wrote to tally tag and attribute use in thousands of web sites made it clear that HTML is a presentation markup language. It may have been conceived for structure markup, but it's used almost exclusively to decorate 2D graphical browser presentation. 

Bob Stein, "graphic design basics"

Part III - The 4 Myth Reasons

While we're at it, let's drop the outdated concept of using tables for field formatting and agree on a straightforward way to do pixel-precise formatting of text, graphics, and interface elements. Such a practice may go against the original ideology of separating a web document's content from its appearance online, but it's far more in keeping with the design goals of many of the web's content providers. 

Peter Bickford, principal of Human Computing

The Great Internet Fashion Show

The <table> element was designed to mark-up tabular data (such as you'd find in a spreadsheet). 

All My FAQs Wiki: Tableless layouts

At face value, this is the most stupid thing ever said about the web. The web is the greatest desk-top publishing system ever!

But, to be fair, it is referring to "DTP" as an esoteric term for the systems for producing high-quality paper-based documents, not to the concept of "desktop publishing" in general.

  • It appears to be a way for people "in the know" to convey certain messages, such as the futility of trying to achieve on the web the sort of precision and consistency that DTP can easily achieve.
  • It is also used to criticise any attempt to achieve presentation control via techniques that are not sympathetic to the nature of the web.

I believe that the nature of DTP has been exaggerated by those who seek to show that the web is vastly different. Obviously people using DTP produce a final version that would print satisfactorily on their chosen paper size. It would stupid not to! But that doesn't mean they always had that paper size in mind. If the only paper size available to them had been something else, that is what they would have exploited. What they often had in mind was a layout concept at some higher level, which the final few percent of the authoring process (workflow) mapped onto a specific paper size.

Perhaps this saying is a criticism of the authoring tools used by some people developing for the web. Perhaps people think that DTP means you "pull at the corners", "use header levels for presentation purposes", or whatever. But ....


This was a DTP package that, as far as I can tell, achieved a vastly cleaner separation of content & presentation than has ever been achieved on the web. If the web was DTP, and DTP meant Ventura, we would not be in the middle of a holy war. Look at Ventura, and weep!

"The web is not paper"

At face value, this isn't far behind the above saying for stupidity!

But, in fact, the message it tries to convey appears to be a similar one. A particular piece of paper has a fixed size. This saying is (I believe) really advice not to publish for a fixed-size rendering. I suspect that many people, like myself, don't use word processors with a specific paper-size in mind. I expect what I write to appear on at least A4 and American Letter size. But in fact lots of what I write with a word processor ends up on the web, as well as or instead of paper.

"Word processors are not paper"!

Some people want us to believe that the web is so unlike anything that has gone before that we should stop thinking in the old ways. But those people ignore the probability that what we did on paper wasn't because it was paper. It was because the writers & readers were human beings. Or at least human beings of "western" culture. We scan left to right. We scan from top to bottom. We focus in from the large to the small then concentrate on the small. We carry position-expectations with us.

Left-to-right and top-to-bottom layout is vital to us. Cats, and probably human beings, have "horizontal" and "vertical" hardwired into the brain. Western cultures tend to use horizontally-aligned text. Some oriental cultures use vertical script. But does any successful culture use (say) diagonal or zigzag script? Few if any.

"Don't expect to be pixel-perfect"

Why not? Some of the web's best brains are trying to develop CSS to achieve pixel-perfection! Those values of the form { someproperty: 5px } didn't get into the specification by accident!

This saying appears to be used in 3 different ways:

  1. It is sometimes a very sensible and valuable reminder that the final rendering is under the control of the user's system.
  2. It is sometimes an implicit admission that CSS positioning is an incompetent system, often used to admonish an author for expecting competence.
  3. It is sometimes used as a strawman, to slap down someone who didn't actually ask for such precision, but instead asked for something the slapper disapproves of.

Take your pick, and act accordingly.

"... tabular data (such as you'd find in a spreadsheet)"

All My FAQs Wiki: Tableless layouts says: "The <table> element was designed to mark-up tabular data (such as you'd find in a spreadsheet)".

Who says? And what is "tabular data" anyway?

Excel (spreadsheet) cells can hold data or formulae. Formulae can (and typically do) refer to other cells. Excel executes any formulae, and presents the resultant data, in a chosen format. (I have a couple of spreadsheets for people to download if they want). HTML tables are very different from this! Yes, there is a common subset, for example simple text or numbers. But so what? What would be gained by restricting HTML pages to such a small subset? (There was in fact a suggestion for adding some spreadsheet-like capability to CSS. But W3C has never suggested that tables should only be used for spreadsheet-like data).

The <table> element was designed to display simple and complex material in rows and columns. See November 1993 in "A brief history of tables". That included "text, multiple paragraphs, lists and headers", where text includes links and images.

Here is what W3C appears to thinks "data" is in the context of tables. See December 1999 in "A brief history of tables". "The HTML table model allows authors to arrange data -- text, preformatted text, images, links, forms, form fields, other tables, etc. -- into rows and columns of cells". That doesn't resemble "... such as you'd find in a spreadsheet"!

The web, and forums of various kinds, show many attempts at defining "tabular data". There is no consensus. Why debate the definition anyway? Why not simply go back to the specification, which includes "text, preformatted text, images, links, forms, form fields, other tables, etc"? With heterogeneous distributed computing, the interface specification has to be the final arbiter, not personal opinions about what "should" be the case. (A second important factor is whether the products meet their quality criteria).

"Tables were not intended for layout"

Who says? And so what?

Condoms were not intended to be placed over gun barrels to keep the sand out. Nylon stockings were not intended to be used as diffusers to soften photographic images. The Romans didn't intend their roads in the UK to be used by motorised vehicles. If the Romans could have imagined motorised vehicles, would they have stated that their roads should not be used for them? I suspect they would have happy with the idea, as long as the roads were not damaged. Sometimes, the fact that something is not specifically included does not mean there was an intention to exclude it.

Should we stop doing all of those things? Only if they are harming people. Remember to remove the gun before using the condom as intended. Remember to remove the woman's leg from the stocking before using it for photography. Before using the stocking as a diffuser, that is. It is OK to use a woman's leg for photography, even though legs were not intended for that purpose.

But, anyway, is it true that tables were not intended for layout? No! Why do the tags and attributes include "row" and "col"? Tables were explicitly intended to layout complex information in a rectilinear (horizontal plus vertical) formation on a display device. All early descriptions of tables were accompanied by such a visual formatting proposal or assumption, with unambiguous accompanying text. For more than a decade the proposed and specified contents of table cells have permitted complex material such as multiple headers, paragraphs, and lists, as well as simpler data. How did all of these things happen if tables were not intended to display complex content in a grid formation?

Laying out such complex material may be subtly different from "page layout", although that isn't obvious. It appears that a decade ago no one really addressed the topic of page layout at the gross level. I'm not convinced that anything was deliberately intended for that type of page layout! Not tables, certainly not CSS1, and not even CSS2. It appears to have been an unforeseen need that became satisfied by the best means available, which was by exploiting the flexibility that those clever designers had built into the table-proposal. (Unforeseen by most of the key players, perhaps, but not all. There was a proposal for a "column" stylesheet feature for layout purposes, made in June 1993). Should designers and authors therefore not have attempted any complex page layout from 1995 onwards, until the "correct" techniques became established globally? Human beings simply don't behave like that!

The web has developed its own nature. It is no longer controlled by the intentions of its parents. It is as pointless to moan that tables were not intended for page layout as it is to moan that your daughter was intended to become a doctor or lawyer or academic, not the actor or artist she is trying to become. Where it goes next depends on the dynamics of publishers, users, development tool & browsers developers, and new ideas, with nudges from standards bodies and governments. It doesn't depend on original intentions! They are now irrelevant.