Category Archives: ideas

i know why search is broken

Search is broken. Google, Yahoo, Ask, Alta-Vista, and on, and on the list goes.

Hundreds of companies, thousands of individuals. I know why search is broken, and I know what needs to be fixed. Now to figure out the how of fixing.

When you’re looking for information, you search on keywords. Google’s been nice enough to rank results by ‘popularity’ (yeah, it’s called PageRank, and it’s proprietary, but it’s a popularity/relevance ranking). The problem is that you have to know what keywords were used. Some places are nice enough to suggest spelling fixes (it’s not ‘brittany spears’, it’s ‘britney spears’).

But that’s not the issue. The issue is that you don’t know what word, term, or phrase to look for. You have the concept you need to find, like ‘module’. Except you don’t think of that word, you think of ‘chunk’. Bam! You’re out of luck: no author would use the word ‘chunk’ when they mean ‘module’, right?

To fix search, we need to search on not just the keyword, but the concept. In English, you’d use a thesaurus.

So, you’re thinking: “This is easy! I’ll just build a comparator that looks at the keyword and then goes through an index of a thesaurus and finds stuff. And we’ll all be rich!”

Hold it, buster. You missed something. This is a perfectly valid English sentence, and you can figure out what I’m saying, too: “Bring me the cooler cooler cooler from the cooler’s cooler.” Cooler is used five times, with the following meanings (at least): hip, less warm, box to keep things cool, jail cell, big refrigerator.

That’s the problem with trying to fix search. Words can mean far too many things in English. But here’s your big chance to figure out a solution: I’ve told you the problem, and I’ve given you the target.

Now go make it work.

a flying ‘bus’

Public transit is an interesting concept, and as I’ve written about before, is not a panacea.

However, I’m wondering if it might be reasonable to apply the concept to flying. Bear with me a minute here.

Let’s say you fly – maybe a lot, maybe not a lot, but when you do, you pretty much go to the same places. Airlines have frequent flyer programs. Why not have frequent flyer cards upon which you not only earned miles, but also could buy flight segments? For example, let’s say I fly a lot between a given pair of cities (maybe Raleigh and Baltimore). A company like Southwest could sell me en masse a bunch of flight segments (good for any segment they fly from any city to any other), and I could redeem them for actual trips.

Taking my example, let’s say I fly a lot between Raleigh and Baltimore. Instead of paying $50 each way (or whatever the current price is) – I could buy 10 flight segments for, say $475. All taxes and fees included already. Then, if i need to fly to Albany for some reason, instead of redeeming 1 flight segment each way, I cash-in two each way, and voila: I have a round trip from Raleigh to Albany.

By selling flight segments in batches, Southwest could guarantee passengers (or at least revenue). The drawback, of course, is that you can only redeem segments for flights that have open seats left.

What do you think about this?

the future of interfaces

Apple recently introduced a convergent device that is a media player (iPod), cell phone, wifi device, and widget player (mini OS X) – I’m going to call it the iCon (since the current name is under trademark dispute). There’s one whole walloping button. Everything else is done via hand motions. Want to zoom in on an image? Grab it in two places, and slide your finger apart. Want to view a photo or video in quasi-letterbox format? Rotate the iCon 90 degrees and the image magically rotates to maintain viewability.

I wrote an article about the brokenness of current GUIs several months ago, and it was published by the ACM’s Ubiquity. You may find it interesting; you may not. You may want to read it in conjunction with this post – either before or after, doesn’t matter to me.

I wonder, though, where we’re actually going in interface engineering. We’ve been stuck on a handful of very limiting techniques for a very long time (command-line and windowing environments). Why are more interfaces not tried? Why don’t we bother to expand our horizons? I think it’s because we’re afraid to experiment.

Interface design needs a lot of work. But more importantly, it needs people who are willing to try anything at least once.

it’s not my problem

Security, air quality, water potability, land use, and the list goes on and on. When any one thing is too big to be one person’s problem, it becomes a problem for the populace, and once it’s everyone’s problem, it’s no one’s problem.

Securing airports is too complicated for one person to do, so a committee tries to do it, and we get security theater (to quote Bruce Schneier). Security should be everyone’s concern, and though it appears to be, we relegate it off to some government agency to handle for us. And then, when they do stupid stuff, we bitch and moan, or just suffer in silence, or sometimes we cheer because we don’t know any better.

Voting is not any one person’s problem – it’s an issue which every registered voter in America should have high on their priority list to do every year. I will be going to the polls in Mebane NC on 7 Nov 2006 to vote. Because if I vote, I exercise my right to have a say in how my government works. Those citizens who elect to not go to the polls to vote lose their chance to have a say in their government. If you don’t vote, you relegate your thoughts, opinions, judgments, and preferences off onto other people who, by definition of ‘opinion’ and ‘preference’, will not always think and act like you.

People who don’t vote are like people who toss candy wrappers out the windows of their cars on the highway. They figure a wrapper or two, here and there won’t affect anyone. And they’re right – to a point. A handful of candy wrappers, cigarette butts, and napkins won’t affect anyone. But it’s not just the candy wrapper they chuck out the window, it the other 14 million drivers who think the same way, and then bitch and moan over how gross the highways look.

Voting and being disappointed with the results does allow you to complain about the outcome – at least you threw your ballot and tried to make your preference win. But if you don’t vote, you can’t complain about the outcome because you didn’t get off your lazy butt and go to the polls. In fact, you don’t even have to get off your lazy butt to vote – you can request an absentee ballot, and vote by mail.

I think we should adopt a policy in this country similar to the one the Australians use – voting should be required if you are registered. Australia consistently has >90% voter turn-out. They fine people who don’t vote.

A lot of people in America are unhappy with their government officials, but an awful lot of them won’t show up on Election Day to try to change the situation. I don’t actually care whether you are a Republican, Democrat, Libertarian, Socialist, Communist, Green, or Independence party member. I don’t care if you’re affiliated with a political party – if you’re an adult in the United States and you’re registered to vote, you should be at the polls on Election Day.

Voting is a privilege in America that many nations’ citizens do not have. Don’t squander your advantage. Unless, of course, you’re content to let other people determine your nation’s course.

a kinder, gentler HRT

As much fun as it would be to be a part of the FBI’s Hostage Rescue Team, what I’m thinking about now is a Hurricane Response Team.

Any church, charitable group, or even a business that wanted to engender some positive community goodwill could assemble a disaster team – volunteers and equipment that could be deployed to damaged areas easily, and self-sufficiently.

I envision such a group having a cargo van and a personnel van. Inside the cargo van could be stowed a gas grill, food, camp stoves, propane tanks, chain saws, wrecking bars, work gloves, and other ancillary safety equipment and tools. A 15 passenger van can comfortably accommodate 10-12 people, their extra clothes, beverages, a couple spare fuel cans, and personal items.

What else am I missing?

beware the fury of a patient man

John Dryden was right.

A single, dedicated man is nearly impossible to stop. And one who waits a long time before exacting revenge is even harder – the target never knows that it’s coming.

I took an introduction to terrorism class during the spring 2000 semester at Hudson Valley Community College. The class served as a presentation of the causal factors behind terrorism, and how defenses against it are developed. As part of the class we were broken up into groups and given the task of picking a target, and then installing defenses against threats we thought to be most probable.

My group picked an embassy (using floor plans of a structure in downtown Troy NY). We decided that we could defend against anything other than kamikaze-style missions and nuclear weapons, but that it would be incredibly expensive to do so. In order to provide round-the-clock guards, we decided to station about 30 Marines at the embassy, install 30+ security cameras, maintain an extensive armory, and put metal detectors and X-ray machines at each entrance.

We had planned three routes of escape/attack for our building: the roof, the river (which was conveniently a few feet from the back door), and the front door. While we didn’t decide to station a helicopter permanently at the embassy, we were prepared for up to three to land on the roof and cart away refugees. To beef-up our external defenses, we installed 4 guard towers on the corners of the building, and put a Phalanx CIWS unit on the roof. We also armored the dock doors (which shielded our planned boat from the outside world) so an attack would not be likely to destroy one of our escape routes.

But the real issue we were most worried about was someone who didn’t bring a metallic weapon with him to the embassy, and yet was still dangerous. A rope, ceramic knife, cane, or any of a thousand other items could be easily smuggled inside the embassy, perhaps even in plain view, and there was nothing we could do about it. That dedicated individual, who may be operating under a shroud of not caring about his own life, would be practically impossible to stop – or even identify.

Even after that class, I have spent a lot of free time noting problems with physical security at various locations. The primary point I noted, due to my employer’s proximity, was the airport. I realized that taking-out the entire fuel supply for the airport would be a trivially-simple task for anyone who could shoot well at long ranges, and who acquired a few incendiary or tracer rounds. Positioning himself back a few hundred yards, and firing a handful of incendiary rounds into the large fuel tanks at the airport wouldn’t be hard – especially considering the fact that they are only typically behind chain-link fencing. And if shooting wasn’t his strong suit, he could just rent a moving truck, and slam through the fence into the tanks.

While performing my more-or-less mindless job of cleaning rental cars, I thought about how those tanks could be protected better. I designed several soft-armored ‘curtain’ arrangements made from layers of Kevlar and thin Lexan sheeting which could be suspended around the tanks from simple frames, and would provide a great deal of bullet resistance.

Stopping a kamikaze driver wouldn’t be too hard, either: just put in place some concrete or steel posts/barriers, and any civilian vehicle would be stopped from getting too close. Stopping a suicide pilot would be difficult, if not impossible, but defense has to stop somewhere.

From both my class experience and further personal reading, I am convinced we spend a lot of money on securing the wrong things. Yes, preventing some nut-job from hijacking a plane is a good thing. However, I think it would be a boon to airline security if people with legitimate carry permits were allowed to bring personal firearms onboard – in the passenger cabin, not just the baggage area.

A primary deterrent to crime is the thought that maybe the person about to be attacked will defend himself. In areas where legally carrying weapons is either inhibited or prohibited, criminals have a much easier time than where carrying weapons is allowed or encouraged.

It would seem to me that it would be a more effective use of security dollars to invest in real physical security and intelligence rather than what Bruce Schneier refers to as ‘movie-plot security’. We’ve spent money to make cockpit doors more-or-less invulnerable, but pull 84-year-old grandmothers out of line to be screened more thoroughly. Security is about identifying the most likely threats, and responding to them. It’s not about coming up with a possible attack, and defending against it alone.

When I worked on the embassy protection project, I kept trying to come up with other attacks that the proposed defenses would be able to handle. And if something we were proposing was really only useful against one highly improbable action, it was listed as discardable if it couldn’t be afforded.

We need more people coming up with real security devices, like my proposed curtains, rather than coming up with movie-plot scenarios.

dating sites

I’ve noticed (for a while now) that most dating and matching sites are all pay based. What I want to know is why? Yahoo! Personals, for example, requires a monthly subscription fee to use it for more than browsing. eHarmony is famous for allowing free scanning, but they charge a (seemingly high) monthly rate. also uses a subscription system. There are myriad others out there that all use the same basic concept – allow people to create a profile free, but as soon as they want to really use the service, it’s going to cost them.

It would seem to me that such a service should be able to be built and run based only on non-subscription revenue. Many people make a stable income from such services as Google’s AdSense. On high-traffic sites, that income has been reported to be as high as $100k/month or more. Even if it took some time to get the service noticed and utilized, I think such a service could be constructed and run entirely off ad revenues. A quick survey of available hosting packages from just one company shows dedicated servers with 4TB/month or transfer and 30GB of space for $99/month.

With such space and bandwidth available, and especially with the bonus of total administrative freedom of a dedicated box, I don’t see why someone doesn’t launch a free dating site. Basic requirements would include a demographics survey, space for an ‘about me’ narrative, a search feature, place to store a small collection of pictures, and some way to contact other members (email via scripted page?). I think it should also have a way of verifying members when they register for the site – a combination of captcha and email /text message exchange – to ensure that scripts aren’t registering bogus people.

A simple policy of “we won’t share your information with anyone without your permission, and we do not guarantee the accuracy of any profile on our site – please report suspicious activity to us at…” should absolve the administrators of any legal repurcussions if something doesn’t work out between members. Obviously, a lawyer should be consulted to get precise wording, but I think this has great potential.

Another feature that could be offered would be private or internal messages – like eBay uses. Members could intentionally change their message delivery preferences to have them delivered to their personal e-mail address, but would default to storing them on the server so that you have to login to communicate. Users should be able to block people they do not want to hear from in the future, too.

I’ve been thinking about this a lot recently, and would be interested in building something like this with some help. If anyone reading this is interested in such an endeavour, please contact me.