News for hackers. Stuff that matters.
Re:The Case for More Traffic Roundabouts by saosebastiao on Friday September 18, 2015 @05:25PM (#10242148)
They missed out on the biggest problem with roundabouts: Their worst-case traffic behavior is cancerous. If any single exit is blocked, entry traffic will become blocked in all directions. And if the road grid is dominated by roundabouts, the problem grows in all directions.

It's not a problem when traffic is just congestion. It's hardly a common problem at all. But when it is a problem, such as is the case with a traffic accident that blocks multiple lanes of traffic, it causes full-stop gridlock to grow outward to the point of full paralysis.

The worst traffic jam I have ever been in was not in NYC, Boston, Seattle, or San Francisco, but rather in Idaho Falls, Idaho on a normally empty country road that was running parallel to the road where a drunk driver caused an accident that blocked the roadway. That single accident caused me and hundreds of other cars to sit in the same spot for almost 4 hours.

Re:The Case for More Traffic Roundabouts by anon4 on Friday September 18, 2015 @05:50PM (#10242270)
One more problem is pedestrian crossings. With a traditional intersection, only cars changing direction cross the path of the pedestrian. On a roundabout, when you cross one of the exit arms, your path is constantly crossed by drivers going straight through, doing a right turn, doing a left turn and doing a U-turn. They are all driving at full speed too, rather than just starting from a stop. It is a delicate game where you gently put your foot forward to signal your intent, wait for the incoming traffic to stop for a moment, then dash to the other side. This is generally my experience, but maybe in other places drivers treat pedestrians on zebras differently.
Re:The Case for More Traffic Roundabouts by mrb on Friday September 18, 2015 @06:01PM (#10242316)
"your path is constantly crossed by drivers going straight through, doing a right turn"

It's only crossed by the same amount of drivers than a regular intersection: those going on the same street as you.

"They are all driving at full speed"

Actually no, the curve of the roundabout (for properly designed roundabouts) slows them down. In fact a traditional intersection causes them to drive even faster since many drivers go in a straight line.

"It is a delicate game"

The nice thing about roundabouts is that you only have to check for cars coming from ONE direction, as opposed to a traditional intersection where cars can come from behind you, from your side, or from in front of you. So crossing a roundabout requires less cognitive bandwidth.

Re:The Case for More Traffic Roundabouts by _delirium on Friday September 18, 2015 @06:14PM (#10242360)
At busier intersections I find traffic signals much nicer as a pedestrian, because while the same number of cars do cross your path, they don't cross your path when you have the walk signal, which is when you'd actually be crossing the intersection. It gives you a nice break in traffic rather than having to push your way out and hope traffic stops for you. This is assuming no right turns on red (as in most of Europe), and left turns only on left-turn signal (typical in cities, at least on major roads). In those cases, there is no traffic at all allowed to enter the crosswalk when you have the walk signal, so you have an exclusive protected crossing.

At lower-traffic intersections the difference is less pronounced, partly because it's easier to force your way across even without a signal, and partly because the signals at less busy intersections may not give pedestrians an exclusive period in the cycle anyway (e.g. by allowing left turns on a regular green).

This might be Copenhagen-specific, but in Copenhagen I also find roundabouts make the pedestrian/bicyclist interaction less clear. At regular intersections, bicyclists can't cross your path when you have a walk signal, but at roundabouts they seem to just go through and are less likely than cars to stop preemptively when they see a pedestrian wanting to cross. Roundabouts aren't very common though, so maybe this would be worked out if there were more of them.

Re:The Case for More Traffic Roundabouts by mrb on Friday September 18, 2015 @06:43PM (#10242457)
"I find traffic signals much nicer as a pedestrian"

You may find them nicer, but they are not safer. The data already shows roundabouts are safer. As the article reports: "researchers have found that roundabouts reduce [...] pedestrian/cyclist incidents by 40%"

Re:The Case for More Traffic Roundabouts by _delirium on Friday September 18, 2015 @07:08PM (#10242552)
That study was specifically in the USA, whose signalized intersections are not really pedestrian-friendly to begin with. For example right-on-red is permitted almost everywhere (except NYC). In that context, I could believe that roundabouts are less bad. I would be interested to see how the comparison generalizes, for example a study of roundabout vs. signalized intersections in Denmark or Germany. It's still of course possible my feeling of more safety at signalized intersections in Denmark is wrong.

Sweden takes an interesting approach in combining both. There are a lot of roundabouts, but most of them also have a signal at the entrance. This makes it easy to cross as a pedestrian, because your crosswalk at the roundabout entrance is protected with an exclusive signal cycle. That forces a gap in traffic instead of you having to force the gap by trying to step out. Possibly even safer than either option alone?

Re:The Case for More Traffic Roundabouts by Tomte on Saturday September 19, 2015 @02:56AM (#10243547)
I don't have a study at hand, but in Germany, too, it is generally accepted that roundabouts are safer.
Re:The Case for More Traffic Roundabouts by brc on Friday September 18, 2015 @06:35PM (#10242430)
You're wrong about pedestrians. The crossings are upstream of the roundabout and you only have to deal with two directions of traffic.
Re:The Case for More Traffic Roundabouts by anon4 on Sunday September 20, 2015 @04:17AM (#10246917)
Not the one I have to cross. The crossings are right where the exit/entry roads begin/end. Which makes sense, since that's the natural progression of the sidewalk and there are two bus stops on the roundabout, so people have to cross right there to reach them. A picture is usually worth a thousand words: https://www.google.com/maps/@42.6686834,23.366936,3a,75y,165...
Re:The Case for More Traffic Roundabouts by ileitch on Friday September 18, 2015 @05:47PM (#10242249)
One some larger roundabouts in the UK (and no doubt other European countries), they have traffic lights on the roundabout, between exits (I'm not talking about the entry lights). If one exit is totally blocked, these lights can allow traffic to continue out of other exits.
Re:The Case for More Traffic Roundabouts by mrb on Friday September 18, 2015 @07:22PM (#10242618)
This "biggest problem" also occurs with traditional intersections.

In both cases, traditional or roundabout, drivers are supposed to leave gaps in the traffic to at least let some cars through the intersection.

In both cases, if drivers don't follow that rule gridlock will ensue. In fact, 2-3 cars stuck in the middle of a traditional intersection is often all you need to completely block entry and exit traffic by preventing anyone from going straight or making left turns: https://en.wikipedia.org/wiki/Gridlock

Re:The Case for More Traffic Roundabouts by foobarian on Friday September 18, 2015 @09:58PM (#10243009)
In New York City the rule is "Don't block the box" and it carries stiff penalties.

I always found it funny how easy it would be to DDOS a roundabout. Get 10 friends together and drive around in a circle all day.

Re:The Case for More Traffic Roundabouts by Tomte on Saturday September 19, 2015 @02:58AM (#10243550)
That actually happens: cyclists have found that to be a very effective and infuriating way of protest on several occasions. You can find lots of Youtube videos.
Re:The Case for More Traffic Roundabouts by chc on Friday September 18, 2015 @11:18PM (#10243157)
Couldn't you block any intersection pretty easily with 10 cars dedicated to doing so?
Re:The Case for More Traffic Roundabouts by FatalLogic on Saturday September 19, 2015 @12:47AM (#10243331)
But then you would be clearly breaking the law, unlike someone just driving around and around
Re:The Case for More Traffic Roundabouts by timthorn on Friday September 18, 2015 @06:53PM (#10242490)
> If any single exit is blocked, entry traffic will become blocked in all directions Typically in the UK, if the traffic is this bad, then drivers on the roundabout leave gaps to enable cars going in uncongested directions to get through.
Re:The Case for More Traffic Roundabouts by Marazan on Friday September 18, 2015 @05:41PM (#10242224)
When traffic density is high you don't build roundabouts. Or if you do you then put part time signalling on them.
Re:The Case for More Traffic Roundabouts by ksenzee on Friday September 18, 2015 @05:49PM (#10242259)
But that's just the point. Traffic density should never be high in Idaho Falls. If you can break the grid to the point where people are in gridlock in Idaho Falls, you've done something wrong.