RouseWorld | Wayward Packet

For the Week of 1/8/01

Don't Just Know What is Wrong
Know What Needs to be Right

(On the wrong side of the tracks in Utopia) — It is easy to say that everything is going to Hell or that the people who are running things are idiots. But how are things supposed to not suck some day if all we can do is gripe about how bad everything is? We all need to have a clear vision of what a non-sucky world would be like and courage enough to occasionally give the world a push in what we see is the right direction (for better or worse, eh?). Otherwise we are all just static in a sea of mildly bitter voices.

Deciding what is wrong is only half the battle, you also need to decide what right is.

It is hard to see sometimes, but the world is made by all of us. The decisions we make every day — how we decide to behave, what we decide to do — all count towards some kind of daily score. Whether the world is a good or bad place is our collective responsibility (but exercised individually). As a practical matter, even the definition of 'good' is your choice, and how you live is your vote on that issue each day.

Here is a challenge — have a clear vision of what kind of world you want to live in, then live each day as if that world is already there. Sound like a crazy way to live? Everyone who has ever made a difference in this world (from Mother Theresa to Mussolini) have had a clear vision of what kind of world they wanted and had the same bullheaded determination to make that world real.

For the Week of 1/29/01

Problematic Distinctions
Structural and Visual HTML

(RouseWorld) — I try to keep reasonably up to date on HTML issues, so I'm also reasonably familiar with the continuous arguments over structural versus visual code. If you don't have the slightest idea what I've been on about, you are likely to find this whole thing dull and pointless — which is one thing I wish those who argue about this sort of thing would understand.

This essay represents a few quick notes on the structural/visual debate. I intend to come back later and fill in what are now arguments without a lot of foundation to them.

Following the Rules

From the William Gibson short story The Winter Market:

"You know what your trouble is" he says when we're under the bridge, headed up to Fourth. "You're the kind who always reads the handbook. Anything people build, any kind of technology, it's going to have some specific purpose. It's for doing something that someone already understands. But if it it's new technology, it'll open areas nobody's ever thought of before. You read the manual, man, and you won't play around with it, not the same way. And you get all funny when somebody else uses it to do something you never thought of."

That to me is the primary argument against people who have hardwired ideas about what a web sites are and what they should do. I don't like everything that has appeared on the Web, but a site shouldn't be criticized just because it doesn't follow the rules.

We write for the compiler

I remember following comp.lang.c for a while back when I was trying to learn the language. Every now and again someone would ask a question about building a user interface. Every time the person would be told something like this:

There's an easy way to figure out if you can do a given task in C. Have you ever seen a teletype? Basically it's a glorified typewriter, used to be what people used instead of terminals. If your input or output can't come from (or go to) a device such as this, then you can't do that in C. Worse, if your input or output can't be done a line at a time (i.e. you need to handle individual keystrokes, as they occur), you can't do that in C.

The above quote is strictly speaking accurate but hardly helpful. Hardly anyone writes code just for the ANSI C standard, they write code for their compiler (and the libraries that come with the compiler). And people do write programs in C that draw pictures and play music. To ignore what people actually do with code, to ignore the code's ultimate purpose in preference to some kind of ideal about a language should be for, seems silly to me.

In the same way we create HTML pages and web sites that run on the browsers that we use. I don't write HTML just to satisfy W3C's HTML validator, I write HTML that will be displayed on a web browser. Because the ultimate purpose of an HTML page is to be displayed in a browser, how the page displays is important.

Tools aren't structural

More importantly, computers cannot understand structure enough to enforce structural design. Even under limited conditions — like outlining — the software I've come in contact with makes bad decisions about the structure of the document I'm creating. On this page all of the sub-topic headers had to be marked by myself as H4 titles. Adobe's GoLive is a pretty good visual HTML editor, but it cannot make good decisions about content that it cannot understand — like the English language.

Software tools cannot give HTML structure unless the person who is using them provides the structure.

People aren't structural

Perhaps I shouldn't say that people aren't structural. People look for and appreciate order and patterns, but the key words here is look for. I would argue that half of the look of a prepared paper or presentation is real structure, the rest is design elements to make the paper or presentation more pleasant to read. Even how people break paragraphs or construct sentences is partly a design decision.

Structure is partly visual design.