Archives for September 2006

Google Lifestyle?

Earlier today I had to cancel a dentist appointment. I looked in the file where I normally keep my contact info, but didn’t see the dentist’s phone number. So I searched for [mydentistsname cupertino] and up popped a Onebox with his business address, phone number, and a little Google map:

The dentist in a Onebox

Then a bit later, someone asked me, “Which is bigger, a square pan that’s 8 inches on each side, or a round pan that’s 9 inches across?” I was on the computer, so I opened up a tab on the browser and typed into Google’s search box: pi*(4.5*4.5) as I’m mouthing the words “Let’s see, pi r squared, r is 9 divided by 2 so r is 4.5”. And Google says

63 square inches and change

which means that the square 8″ by 8″ pan is a little bigger at 64 square inches.

Every search engine can do some tricks like this. It’s just interesting that 5-6 years ago, I wouldn’t even think of hitting a search engine for stuff like this. 2-3 years ago, I didn’t realize what a difference it made to have all my email easily searchable. Six months ago, I didn’t know how handy it was to keep my Firefox bookmarks in sync. I suspect that in a few months, I won’t believe how handy it is to have a desktop search that lets me securely search for any webpage I’ve read in the past with just a couple keywords.

So the question is, what is waiting a few more years down the line? Maybe I won’t be able to imagine life before my documents sat in a magic Writely cloud where I could get to them from anywhere? Will cell phones evolve straight to internet-connected computers with an always-on broadband connection? Will people record their whole lives, because storage for audio and video will be so cheap?

What do you think will be the must-have gadgets in five years? What will the killer services be?

GoodKarma radio show and a clarification

Recently Vanessa Fox and I did a GoodKarma web radio show together. The MP3 is here if people want to listen to it. I’d like to clarify one thing though. At one point about 40 minutes in the show, we were talking about some web sites that had been hacked, and I mentioned we hadn’t been able to contact the webhost involved. I meant to cough and half-mention the name “iPowerWeb,” (to try to get someone to contact me) and I said “Oneupweb” instead. It was a pure brain freeze on my part, and I apologize for the confusion. Thanks to Lisa Wehr and the folks at Oneupweb, who did contact me; they gently and humorously reminded me that it’s very easy to contact Oneupweb–they even offered to respond to smoke signals.

On the bright side, we did manage to contact a couple people at iPowerWeb too, and I appreciate them responding.

By the way, if you haven’t completely gotten your GoodKarma fix, Vanessa and Danny did a funny show the week before about how Buffy the Vampire Slayer teaches good SEO; the MP3 is here. Next I’d love to hear what The Princess Bride can teach about SEO.

Talk like a Pirate!

Holy crap, the holidays do sneak up on you, don’t they? The Inquirer reminds us that it’s already September 19th, which is International Talk Like A Pirate Day! Yarr! Shiver me timbers, it be today!

(I’ll look at comments later, and any comments not in Pirate-speak will walk the plank! Arrr! If you have trouble with that, you could buy a Pirate keyboard.)

Have a good day, mateys!

Update: Aye, i had no idea thar war so many pirates on the web and at Google. Har’s a picture o’ someone in the cafeteria who dressed up (showin’ his picture with his permission):

(Yarrr! Not reproducing the picture here because my camera phone has been struck with scurvy and it cannae email pictures. Or the packets have hit a hidden sandbar in Verizon’s network, laddies.)

At Google I also saw pirate T-shirts and (no joke) someone had what looked like a li’e gray parrot on their shoulder, argh!

Hi, I’m a CSS idiot

Hi, I’m a CSS idiot and I have a question, so I’ll ask you. If I have a web page A that includes an iframe B, can page A override the CSS in iframe B?

Thanks in advance. 🙂

Changes in url queries

(Writing this one fast just to do a quick explanation.)

One truism you learn quickly at Google is “you are not a typical user.” If you’re reading this blog, the truism probably applies to you too: you’re much more likely to be a power-user, an SEO, a librarian, or someone else who is familiar with the site: operator or the info: operator. But it’s important to remember that many Google users aren’t like that.

Bearing that fact in mind helps explain a recent change in how we handle url queries. Some people call these navigational queries, but at Google a navigational query is typing in [HP] and expecting to see high in the search results. A url query would be something like [].

Previously we treated the query [] like the query [], and now we treat it like [“”]. The query [] returns the single url if we have it in our index, along with other choices like “see backlinks for” (I’m oversimplifying a little, but nothing too bad). The query [“”] searches for that as a phrase, and thus returns the ten best matching urls, which will usually show at #1 or high in the search results.

Why did we make this change? Bear in mind that you, gentle reader, are not a typical user. If super-duper power-users who know how to refine site: or info: are a set of people we’ll call N, you are probably in that set N. There is a whole different range of people M who just type in to get to, and who sometimes misspell the url. In math, M >> N means that M is much greater than N. That is, there are many more people who casually type in urls to get to those urls (and who sometimes misspell those urls) than there are super power users. So this change helps M. The N power users can just prepend “info:” to get to the old behavior.

Let’s take an example. Suppose the query is []. If you go visit that domain, you get a 404 error. Most likely, what happened was the user was typing [] and accidentally left out the “p” from myspace. Which behavior is more helpful to that user? Here’s []:

The info: results

Not that helpful to a user who misspelled the url or who was in a hurry and left out a letter. Now here is the new behavior for []:

The quoted results

Notice that the new behavior gives a suggested spelling correction, results from, and a suggested query correction with another chance to find The second set of results is more likely to be helpful.

We can always tweak/refine/revert if we see problems, but this change is beneficial for many of our users, and expert users can still get the old behavior by just prepending an “info:” to their url queries.