Hacker News Reader

Comment by teilo | Hacker News Reader

"In 2004 the New York Times reported that more than 2,500 of the 3,250 "walk" buttons in New York intersections do nothing. "The city deactivated most of the pedestrian buttons long ago with the emergence of computer-controlled traffic signals, even as an unwitting public continued to push on."

And what do you expect them to do? Since only 2/3rds do nothing, how is the pedestrian to know which 2/3rds that is? They have no choice but to always push the button.

View on HN

What I find odd is the crosswalks here in DC where there's sign that says "Pedistrian MUST push button to cross." Yet it's not true. The button does not appear to do anything and a walk signal definitely comes up even if you don't push it.

Existential. Push button anyway. The other way lies nihilism.

No.. no... don't push the button. It's the only way not to be sure.

Have you tested to see if the delay is diminished? There are so few of those I can't imagine they're placebos. When I lived at GWU I can only remember the ones across Pennsylvania, and everything else was automatic.

I haven't done controlled experiments, but anecdotally it makes absolutely no difference.

What I find odd is the crosswalks here in DC where there's sign that says "Pedistrian MUST push button to cross." Yet it's not true.

Actually, it is true at some of them. If you don't press the walk button at some of them, the walk light doesn't turn on and the light is on red for a shorter period of time.

Another way to interpret the disabled walk buttons is to say that pushing the button N times is N times as effective as pushing it only once!

Why don't they remove the inactive buttons?

It's probably more expensive than just snipping the wires.

Why would they even bother to do that? Just deactivate the buttons in software.

I don't think those buttons were driven by anything we would consider "software."

Do you mind explaining? I can't tell if this is meant as a joke or a serious counter-argument. At any rate I assume there is some sort of firmware controlling the lights [edit: guess not, read juson's comment below], monitoring the buttons/ induction loops, etc. and the easiest way (to my mind) to disable to buttons would be to disable them in this software and have the system stop monitoring them. However if you have any insight into how the lights work that would be great.

hey kgermino,

Sorry to butt into the convo, but I was a summer intern at a major "traffic light" operation (for lack of a better term) in the southeast (aka, they did all of the traffic lights and systems for the southeast USA). It will seem almost shocking to many here on HN, but the traffic control industry is lagging behind drastically.

I believe the newer systems started going IP based just a few years ago, and (gasp) wireless networking of traffic lights is a really new thing. It has to do with the environment the system must handle and how "tamper-proof" it must be, which leads to hardware and software being waaayy behind the curve.

The main point though, is that it would be very reasonable to think that some hardware, especially if installed before the 90s, wouldn't have "software" to just "turn off monitoring". Even the newer boxes aren't completely controlled by software, they have physical circuitry that operates as a "fall-back" to make sure that things like two green lights on adjacent roads can't be green at the same time, even if the software was hacked, had a bug, etc.

Interesting. For whatever reason it didn't even occur to me that traffic lights would still have a "hard-wired" circuit available to operate the lights. I'm a Computer/Electrical but I tend to think in terms of coding rather than being able to solve problems with circuits. But that's interesting and makes me think I should look into an internship/coop in that area. Thanks

Even moreso than that, many of the lights in New York still run on genuine relays. You can easily hear them at many street corners in Manhattan, happily thunking away. (The New York Subway is also still powered almost exclusively by relays, even to do complex things like monitor train speed. It's both impressive and disturbing to see what they were able to accomplish a hundred years ago without modern computing hardware.)

Any recommended links on where one could read more on the NY Subway systems you mention? I'm fascinated by pre-electronic computing, and this sounds like a particularly interesting example.

It's still insulting.

I'd be far more insulted if the city spent $2 million [Made up Number] it doesn't have removing buttons that did no harm to anyone.

Why not spend almost nothing to put some sort of notice on the button that it's no longer useful?

"almost nothing" would still equate to sever thousand dollars, probably over $100k. Not to mention future costs, a sticker would probably need to be replaced after a year at most, and a sign would cost significantly more. If you decide to just put them up and let them be there's a high likelihood of vandalism or damage that isn't attractive. All for signs that many people probably wouldn't notice. Again it just doesn't seem that letting people know the buttons do nothing is worth the money that it costs.

The buttons are almost certainly a disease vector. I'd hardly say they aren't causing harm.

Well, because the point of the article(s) was that the user feels happier if they feel they have control. Besides which people would complain if they physically removed the buttons. Leaving them inactive means nobody is the wiser, and everyone gets on with their lives.

If someone pushes the button and the door doesn't close immediately I wouldn't say nobody is the wiser. More likely, someone just caught the common cold for no reason.

Removing 2,500 or more buttons throughout the city would require a lot of man-hours, transporting, storing and disposing of the old buttons. There would also be wiring to remove or block off, cover plates to be made, distributed and installed. And hopefully some sort of communication to the entire population of the city including visitors to explain why the buttons are going away.

It says in one of the linked articles it would cost roughly $400 per unit. That's $1,000,000 right there.

Because having a button there, even if it isn't active, reduces the number of people who will walk when they shouldn't, because they think that the light will change sooner because they pushed the button.

No, it increases the number of jaywalkers. If you know that the light is on a timer and you will have your turn in 15 seconds, you have little reason not to wait the 15 seconds. But if you have to press a button for something, you deserve quick response to the button push, and if it doesn't come quickly, you give up in disgust and cross against the light.

Even when those buttons DO work they aren't instantaneous. You presume a world where pedestrians have almost no patience. A more likely scenario is that people start pushing the button more and more trying to get it to work before they "give up in disgust".

Reply on news.ycombinator.com