It’s dangerous to go alone

March 14, 2002

The Fighting Whities

It seems that an intramural basketball team has named itself “The Fighting Whities” in an attempt to raise awareness of stereotyping and, one assumes, to protest the various sports teams using Native American names and imagery. Their jerseys feature a 1950’s-style drawing of a white man and the phrase “Every thang’s gonna be all white!”

I predict that this won’t be especially successful. Looking through the reader comments, you see of white men making comments like “So what?”, “This is funny”, “Where can I get a jersey?”, and “Since I am not offended, Indians should not be upset about the Redskins” (not actual quotes).

The argument made here is: white people don’t get offended when people make fun of them, so why should anyone else be offended? The problem here is that the circumstances are not symmetric. Isaac Asimov, in the chapter on ethic humor in his Treasury of Humor, writes:

The line between funny satire and outright offense is almost vanishingly small, and it exists in different places for the person who is of the ethnic group being satirized and the person who is not; for the person of an ethnic group who feels secure because he is surrounded by others of his group, and for the same person who feels alone because he is not; for the person of an ethnic group who is feeling good that day, and for the same person who has been unsettled by the morning headlines.

The issue of power and security is undoubtedly a big part of it. White men, as a group, are well represented (some might say over-represented) in positions of authority and prestige. The same cannot be said for other genders and ethnic groups. Hence, white men are secure enough that they can laugh off jokes and not worry about slurs.

Some argue that using Native American names and imagery for sports teams honors their legacy, but this is a weak argument at best. If the people supposedly being honored are in fact offended, you don’t get to tell them they’re wrong. Doing so only reveals that their opinion does not, in fact, matter—which would not be the case if honor was the true intention. (found at Blogdex) #

All your reference are belong to us

In my recent post about Google bombing, I was one of many, many people who referenced an article discussing how weblogs disproportionately affect Google. The speed and scale of the response has prompted another article, which examines how ideas spread through weblogs.

Quick-spreading ideas are hardly unusual on the Internet (although the ideas themselves can be quite unusual). The difference with weblog-borne ideas is that they can often be traced. The time-based nature of weblog posts make it possible to refer to things in the past and be reasonably certain that the reference will still be valid in the future.

Indirectly, this argues that it’s more useful to refer to specific weblog posts than to the weblog itself when passing on links. #

Bad URL, no cookie

John Rhodes, of WebWord, has posted an article about the causes of linkrot. While there’s been a few articles on the subject before, Mr Rhodes is the first I’ve seen to blame Content Management Systems for a lot of the problems.

Large, professional web sites store documents in databases and use scripts to combine the document’s content with the site’s navigational interface and advertisements. That’s fine and dandy, except that the systems used often create long, impenetrable addresses. That’s also okay, since most people won’t be typing the addresses in directly. The problem is that every CMS has a different way of generating addresses. That means that switching the CMS breaks all the links into the site.

While it’s probably impossible to do away with linkrot entirely, I don’t think Mr Rhodes’s suggestion of having sites notify each other when addresses change will scale that well. Better to try to keep URLs unchanging.

The most obvious strategy to employ is that of information hiding. Ideally, your URL reveal nothing about your particular server configuration. It’s fairly simple to hide simple queries. Some simple server tweaking allows you to transform, say “http://www.greeleytrib.com/article.php?sid=7262” into “http://www.greeleytrib.com/article/7262”. Note that the second (purely hypothetical) address could continue working even if the site stopped using PHP.

It may also be useful to hide the structure of sites by using directories instead of files. That is, instead of putting site information at “http://www.example.org/about.asp”, use “http://www.example.org/about/”. Again, this hides the implementation of the about page, allowing behind-the-scenes changes, but it also allows the about page to expand into multiple pages without much confusion.

I say “may”, because there really is no distinction between directories and files on the web. If I had a server configuration to play around with, I’d look into eliminating trailing slashes altogether. Officially, “http://www.example.org/examples” and “http://www.example.org/examples/” are different addresses, but actually making a distinction between the two would only serve to confuse people. I say, use the one without a final slash. It’s shorter and easier to read aloud. Pages within the “examples” hierarchy would still involve the slash, as in “http://www.example.org/examples/specificexample”.

The problem is that most web servers would prefer to use the address with the slash, since “examples” would probably be implemented as a directory in the server’s file system.

Note also that those example addresses don’t include “.html” or “.php” or similar extensions at the end. Again, those are irrelevant. Unless you’re specifically referring to an HTML version of a document (as opposed, say, to a PDF version), there is no need to note the file type in the address.

Hiding implementation details is only one technique for creating durable addresses, but if things are otherwise well planned, it can be the most effective. #