a qmnf by any other name

By Todd Stadler · Thursday, June 5, 2003 1:39am

You know what? The Web's name space is getting a bit polluted.

I realized this when trying to find a random gibberish URL through the keyword search at whois.net.

I expected that, by randomly (okay, all you computer and math geeks out there, stop breathing so hard — I realize it's not really random) pounding keys, I might find one or two domains that contained the resultant gibberish sequence.

So like any good typist, I happened to pick a sequence from the home row: lkjs. I was somewhat surprised to find that seventeen domains contain this sequence, only one of which — the Norfolk Jin Sei Kai Karate club — appears to not be a gibberish URL.

I was intrigued. Have all the good names been taken that domain name squatters have started registering random keystrokes? Or are these seemingly disposable domains being used for more nefarious purposes?

Apparently, not all of them are. Amazingly, myblahsdfkjkllsjdflksjlksjdflkjsldkfjasdf.com seems to be the URL for Tucson Airport Transport. Why a legitimate business would place their Web site at that address is beyond me.

But it got me to thinking. I figured that most three-letter domains are taken, due to the preponderance of TLAs in business names.

So I set off to find a four-letter sequence (no numbers or punctuation) that did not occur in a registered domain name. Not necessarily an active Web site, mind you, just a domain name somebody bothered to reserve.

My first attempts were rather unsuccessful. Running my fingers across the home row to produce sequences like asdf resulted in such wonderful domains as aasdfksdfsieu.com, a9s7df09as7df098asdf.com, aboutasdfasdfasf.net, and afdasfsadfasdfas.biz (the latter two of which sound like the kind of domain you get when your first choice was already taken). And, of course, the potato and potahto of the let's-hit-purt-near-every-key-on-the-keyboard URL, 1234567890qwertyuiopasdfghjklzxcvbnm.com and 1234567890-qwertyuiopasdfghjklzxcvbnm.com.

Continuing to play around, I next searched for registered domains containing the sequence "aaaa".

I expected to find a few domains with such a sequence, much as every phone book seems to have businesses named something like "AAAAA Aardvark Auto Repair", although the benefit is far more obvious in an ordered, static phone book.

Still, I was surprised to find such memorable domains as 1111111111aaaaaaaaaa.com, 11aaaaaepfhk.com, 1700-1900aaaantiques.com, and a-aaaa.com (whose motto must be "the hyphen stands for quality").

In an attempt to find a less popular sequence, I opted for a less popular letter, and let loose the full improbable fury of "qqqq" on whois.net's search engine.

But clearly, the same logic that led me to believe few to no domains would contain that sequence led other, perhaps more clever, people to register domains with it, perhaps in the hope of drawing in traffic from all the people searching for domain names containing "qqqq". Brilliant!

Thus, we have, more or less in order of my incredulity, aqqqq111.org, maiqqqqq.com, qqqqcasino.com, www.qqqqqqqqqq.com, www.qqqqqqqqqqq.com, www.qqqqqqqqqqqqq.com, www.qqqqqqqqqqqqqq.com, www.qqqqqqqqqqqqqqq.com, www.qqqqqqqqqqqqqqqqqqqqoiuyoiuyoi.com, and, just to prove it can get sillier, qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqq qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqq.com

Clearly, I wasn't going to find an unused sequence by repeating a letter, no matter how unpopular that letter may be. The allure of repetition must have been known by all the shrewd businesspeople who registered these domains!

So the kid gloves came off. I thought of really improbable four-letter sequences such as "qpoi" and "qnwm", but those showed up somewhere in the domain name space (the latter found in zsqnwmh.com, a name apparently cobbled together from the anglicized initials of its Chinese owners; leave it to the Chinese to ruin my fun).

I got more and more desperate, mumbling to myself, "surely no one has ever registered a domain with the unholy sequence "qjqj" (qjqj.com and qjqj.net) or "jqjq" (bq--3bigkxvxtdpvjqjqxeymqmfc.com, which, while not as extreme as the one with sixty-three q's in it, holds a special place in my heart for being most untypable).

And I don't even want to talk about "qjxj" (cqjxjt.com and qjxjy.com).

Finally, I decided it was contest time. Who could be the first person to find a four-letter sequence that did not appear in whois.net's list of registered domains?

I was all stoked, because everybody likes a contest.

But I decided that before I had to find something in my house to give away to one of my ten friends who reads this page, I had better do some research.

And by research, I mean "ill-conceived, brute-force programming".

I banged together a PHP script to query whois.net, starting at "aaaa" and, in theory, going all the way to "zzzz", some 450,000 queries later.

Except that, surprise, whois.net didn't like being incessantly queried like that, and so shut off access to my dumb little script.

Either that, or I really bungled the programming. It's so hard to tell the difference.

Either way, I managed to search for domains containing patterns "aaaa" through "aasp", an appreciable 485 queries, but still only one-tenth of a percent of all the possible sequences.

It really makes you stop and think about exponents, doesn't it?

Anyhow, of those 485 sequences, only six of them were not found anywhere in whois.net's database of registered domains: "aajq", "aajx", "aakq", "aaqd", "aaqf", and "aaqj".

Yes, eager domain name squatters, now is your chance! Run, run and register aajx.net or aaqdaaqfaaqj.aero or whatever it is you think will be resold for millions of dollars.

So 1.2 percent of the first 485 four-letter sequences are apparently not used in any registered domain names.

I have to imagine the percentage for the whole four-letter-sequence space would be a bit lower, as "a" is a fairly common letter.

But it all goes to prove what I said at first: the Web's name space is getting a bit polluted.

Oh, and that people are stupid.

But that goes without saying. Around here at least.