Hacker News new | past | comments | ask | show | jobs | submit | more zephjc's comments login

It's important to make a semantic distinction between 'streets' and 'roads'. A typical one is:

Streets are things which are where humand can play, people bike and walk, go to stores, around the places we live and work, etc. They are 'Places' where people can live about their day to day lives. Very human-friendly.

'Roads' which are higher-speed connectors between Places. Highways are a very high-speed type of roads.

In a typologically healthy region, there is a clear distinction between streets in roads, but in the US there is a blurring, and we often see what are sometimes dubbed 'stroads' - a mutant street-road hybrid, the sort of thing we typically see in the US with wide, fast, multi-lane streets, lined with strip malls and the like. They're hostile to human and make walking between locations at best boring and tedious, and at worst dangerous.


This is interesting stuff. Where'd you learn about this distinction?


Strongtowns I think was my first introduction to it. Here's a good recent article on it

http://www.strongtowns.org/journal/2016/5/22/engineers-shoul...

Note his definitions are a little different than what I used, but have roughly the same point


Look to how Tokyo does things. A few takeaways:

- mixed use zoning: reduce requirement for long trips by mixing many compatible types of commercial with residential, and remove single housing type developments. Also, allow many smaller apartment complexes mixed with single family houses, instead of segregating housing types.

- street design: No more hierarchical/dendritic street layouts. That is, no more dead-end streets, which lead to collectors, which lead to arterials - you're bottlenecking a huge population through a very small, fast, and unsafe road system. Instead, make the streets highly connected, and narrower to encourage slower but steadier car traffic, and blocks shorter. Porous streets networks can route around bottlenecks and can have many more concurrent cars than even crazy-huge Texas-style freeways.

- no big street setback requirements: encourage density by removing crazy suburban-style setbacks.

Edit: I wanted to make a plug for form-based zoning, which is zoning where the form (building type) is zoned, not its use. This doesn't necessarily refer to its style (Neoclassical, Modernist, etc) but how it interacts with the surrounding buildings on the street. E.g. buildings above a certain size might not be allowed in an area, and not be allowed to take more than N number of yards of street frontage. Setbacks of a certain size might be prohibited, or allowed. This allows of a reasonable number of mixed uses like restaurants, shops, and other day to day commercial uses to coexist with residential. This does not mean that heavy/noxious industry can be built up there. This was the error the Euclid v. Ambler decision made 100 years ago: they threw the baby out with the bathwater by restricting zoning by type; there is not a small amount of racism that came with Euclidean/exclusive zoning, e.g. removing a formerly viable way for immigrants to start a business with a house above (a la Bob's Burgers) and thereby increasing the barrier for success.


> - street design: No more hierarchical/dendritic street layouts. That is, no more dead-end streets, which lead to collectors, which lead to arterials - you're bottlenecking a huge population through a very small, fast, and unsafe road system. Instead, make the streets highly connected, and narrower to encourage slower but steadier car traffic, and blocks shorter. Porous streets networks can route around bottlenecks and can have many more concurrent cars than even crazy-huge Texas-style freeways.

Alternatively, you COULD use dead-end streets...that are only dead ends to cars, but have pass-throughs for walking and biking. For an extreme example, see Houten, NL: https://www.google.com/maps/place/Houten,+Netherlands/@52.03...

http://www.citylab.com/cityfixer/2015/06/a-case-study-in-bik...


Those are pretty recent designs, but I like them for the most part. I agree, as long as you are not hampering human movement, keep car traffic slow and safe for said human movement, networking with bike paths is fine in my book. In an existing dendritic US-style street system, retrofitting with bike paths between streets is a really cheap and easy way to encourage biking (the hardest part is probably dealing with existing property rights)


> No more hierarchical/dendritic street layouts.

It's worth playing some Cities:Skylines (which has a fairly accurate traffic simulator, particularly with the Traffic++ mod) to understand how the road hierarchy came into being. Or, for that matter, trying to drive through a grid-based city like Manhattan or SF.

You get very large traffic jams. The problem is intersections, and particularly intersections where traffic backs up to the previous intersection. When this happens, a traffic jam tends to spread across the whole city; incoming traffic can't clear the bottleneck fast enough, so the bottleneck just grows like a cancer until it envelopes a whole neighborhood.

Oftentimes, the solution to a traffic problem is simply to bulldoze a few intersections. By doing this, you give cars a buffer. It increases the median trip length but it also increases vehicle speed and road throughput by more. It turns out that the major contributor to traffic jams is the acceleration of having to start/stop at traffic lights and when turning.

Self-driving cars (or just ubiquitous turn-by-turn navigation) could change this equation by intelligently routing cars around bottlenecks and avoiding the neighborhood entirely, but as long as drivers have imperfect information about traffic conditions and tend to take the shortest route to their destination, this will remain a problem.

(I've had great success with using pedestrian paths to provide cut-throughs between dead-ends and nearby intersections, though. And with providing pedestrian paths under or over those intersections so that people don't have to wait for stoplights to cross the street and don't stop traffic with their jaywalking. The game unfortunately has pretty terrible pathfinding for pedestrians and won't let you build compact staircases, so this limits their usefulness to real problem intersections, but in real life I think many suburban cities could drastically improve their walkability/bikeability just by building raised pedestrian footbridges over their major arterials.)


I really enjoy C:S, but even with Traffic++ the traffic simulation is pretty wonky and shouldn't be taken as reflective of reality.

But here's some advice if you're having trouble with traffic jams in your grid systems: use more one-way streets. If you've converted your city over to a 100% one-way grid and still have backups, you probably need to work on your mass transit and freight rail systems. I've made functional cities where every single road was open only to pedestrians, cyclists, and service vehicles. No cars.


According to Jeff Speck (https://www.amazon.com/Walkable-City-Downtown-Save-America/d...) 1 way streets have a negative economic impact for the business that line those streets. Which I think makes intuitive sense. Anyway, that doesn't matter in C:S, but it probably matters in the real world.


Is this assuming all streets are one-way? If they are, than I wouldn't expect an economic impact on businesses. Since the amount of products people buy - and the money they are willing to spend - should stay the same, than the total money influx would remain constant. Therefore, the only change would be how this money is distributed. Since all stores would lie on one-way streets, none would be more impacted than the other.


It's definitely about the specific businesses that are on the one way street. And I might recall (been awhile since I read the book), that he specifically used the example of changing from two way to one way and that the result is businesses on the affected street lost revenue.

It makes intuitive sense. When a street is two way cars pass by both directions each day increasing the likelihood of stopping at a store on that street. When streets are one way, then stores only get people driving by once a day, and so they lose revenue.


Something that also doesn't help with a lot of intersections is dense street parking (blocks line of sight for approaching traffic). It's problematic in cities and getting worse in suburbs. The street I grew up on, was generally empty of parked cars making it a great place to ride or skate in the 1980s. Now, that same street has loads of parked cars - people have second cars, use the street instead of their garage, or their adult children have cars.

Roundabouts would also drastically help many US cities currently relying on all-way stops.


Isn't a large contributing factor to this people being in the intersection ("blocking the box" in NYC) when the light turns red, causing literal gridlock? People, myself included, have a visceral aversion to sitting at a light as it cycles through and not moving, so the instinct is to move forward even if you're going to be sticking into the box a little bit. Then the person behind you does it because they've been waiting just as long as you, and suddenly the perpendicular traffic can't move at all because you're blocking their green light.

I always thought that was why I saw posted fines for blocking the box in NYC but never saw such signs elsewhere (although I've never driven in California).


> Self-driving cars ... could change this equation by intelligently routing cars around bottlenecks and avoiding the neighborhood entirely

Self-driving cars can do even better than that - they can eliminate the "stopping" nature of the intersection entirely. If the self-driving cars are able to determine the position and speed of other cars approaching the intersection, you can just have the traffic streams pass right through each other. [1]

[1] Autonomous Intersection Management: Traffic Control for the Future https://www.youtube.com/watch?v=4pbAI40dK0A


I gotta agree with thescriptkiddie, C:S is an imperfect model of how traffic behaves.

> Self-driving cars (or just ubiquitous turn-by-turn navigation) could change this equation by intelligently routing cars around bottlenecks and avoiding the neighborhood entirely, but as long as drivers have imperfect information about traffic conditions and tend to take the shortest route to their destination, this will remain a problem.

Self-driving cars or drivers with good mapping are still limited where they can go when they have a street hierarchy to deal with, forcing all cars onto the same few arterials.

> Oftentimes, the solution to a traffic problem is simply to bulldoze a few intersections. By doing this, you give cars a buffer. It increases the median trip length but it also increases vehicle speed and road throughput by more.

This is all well and fine in a game, but increasing street speeds kills the street life (figuratively, and sometimes literally). Slower but more constant speeds are better for everyone involved. For walkers, bikers, and even drivers. Ask yourself this: would drivers flip their shit more often when going slow but steady down 15-20 mph hour streets with stop signs (or roundabouts), or when they're stuck at long traffic lights regardless of how many lanes they have?


The benefit of faster speeds isn't driver convenience, it's that you get cars off the road quicker for a given travel distance. Each city has a carrying capacity for the number of cars that may be on its roads at once, which is determined by the length of the road network and number of lanes. They also have a certain number of trips generated, which is determined by population. The number of cars on the road = trips generated * average travel time per trip. When that exceeds the carrying capacity of the city, average travel time increases, which causes a cascading effect that eventually results in gridlock.

The same effect plays out locally, on each individual stretch of road. When integral(# of incoming cars - # of outgoing cars, time) > carrying capacity of road, the road backs up, which increases the time required to traverse it, which further exacerbates the backup. This is why multi-lane arterials can reduce congestion; they can move a lot of cars off a given stretch in a short period of time, and provide a linear buffer where momentary oversupplies can collect without backing up the previous intersection.

You can also see this effect by looking at traffic maps of say, SF (grid layout) vs. Sunnyvale (arterial/collector):

https://www.google.com/maps/@37.7811106,-122.4106957,16z/dat...

https://www.google.com/maps/@37.3724565,-122.0375532,15z/dat...

Both of them have shitty traffic, but in SF the traffic spills away from Market street onto many of the side streets, such that no matter where you go it'll be gridlock. In Sunnyvale, much of the traffic is confined to major arterials like El Camino or Matilda, which are slow but still move, and side-streets that parallel them are often relatively clear.


> The benefit of faster speeds isn't driver convenience, it's that you get cars off the road quicker for a given travel distance.

Yes but that isn't actually what happens because the use of a street or road isn't by one car from A to B but by the continuous use over time across a section of the street.

Say you had a single 1 mile arterial in a city, and it's the only way of getting from one half of the city to the other half. There are few points when a cars are "off" of it (except maybe late at night) - the rest of the time it is a near constant high speed flow.

You're not wrong to say that it gets any given car off the road quicker, but that is if you're focusing on the one driver's trip, as opposed to focusing on the use of and experience of being at that section of road. If it was an old town which had its main street become a high speed arterial, you now have an experience for any pedestrians who might want to use the (probably few remaining) stores along that road be not unlike walking along side a freeway - unfun and dangerous.

By focusing on any one driver's trip experience, and not the street experience, you're essentially damning the street experience for the potential sake of some extra time saved (if its across a city, perhaps on the order of 10 or so minutes).

Of course when you have nothing on the street worth being around (like most of El Camino Real), you want to get passed it ASAP. (SF problems are a whole other hairball of outside commuters plus residents who insist on using cars.)


Did you intend to paste the same quote twice?


I did not, thank you; fixed :)


Tokyo has many intersections with over/underpasses. The 2-4 lanes in the middle go over or under the intersection. The outer lanes connect to the intersection.

That said Tokyo has plenty of traffic.


The lack of on-street parking was something I noticed visiting Japan. Streets can be much smaller when they don't have 20-24 feet of parking. In the US, a massive percentage of our most valuable land goes to subsidized vehicle storage. It's insane.


Narrower streets, wider sidewalks, more trees.


Actually most minor streets in Tokyo have no sidewalk. You can just walk on the street.

Cars drive through slowly, of course.


German suburb showing the example: http://i.imgur.com/DHhjSQ8.jpg


I remember seeing some video on civil engineering twenty years ago where they talking about how that was a major mistake they had learned from. Arteries just cause traffic jams, while permeable neighborhoods allow traffic to diffuse through it.


Re: Street Design

Japan is designed with public transport in mind. Most everywhere in Japan is within train/bus/bike/walk distance. You don't need 4 lane high ways when there are maybe a few dozen cards on any particular stretch of the high way at any given time. I drove from Nara, Nara to Naruto, Tokushima via highway and saw maybe ten other cars on the highway. That's a distance of 174.5km; roughly a 2 and 1/2 hour drive. If I drove for 2 and 1/2 hours on any stretch of highway in California, I'd see ten cars every few minutes.

Less cars in the road, in general, means roads can be more narrow.


For sure. As I understand it a lot of the build-out in Tokyo into what were more rural areas was done by the railroad lines - build a km or two, add a station and more developed land, repeat.


This. You cannot consider Tokyo as an example of (successful) urban planning without understanding the role of the railway companies.


I definitely agree!


Does the perfect city really need to focus on higher population density? It seems like less density is ultimately what most people would want.


yes. nature hybridized with human living. future forest primitive


That's basically the north-of-mainstream thinking in urban planning, right? I certainly agree with you, but I also think it is a wasted opportunity to just build what leading thinkers agree is a good idea.


More or less. Some aspects of Western urban planning are rather big on things like Complete Streets, where you have segregated bike lanes added between sidewalks and street. These are fine for wider streets as seen in many American cities, but smaller streets which force cars to slower speeds don't necessarily bikes and even pedestrians segregated.

E.g. this slice of suburban Tokyo: https://www.google.com/maps/@35.7368825,139.5634333,3a,75y,2... Note the speed limit is 30kph, or around 18mph.

This seems to be the default building style in much of Japan. You will still find larger streets which are arterial in nature, but they're usually still very bikeable and walkable, and still porous to smaller streets like the one linked above.


It's really bad - the underlying pattern of change is tearing down a bunch of buildings an running big highways through the cities.


Not to mention the tens of thousands of people evicted from high density buildings and ramping up suburban sprawl. Will that highway, after deducting maintenance and environmental damage, ever hope to recoup the costs in city services from handling low density housing?


You can also see huge blocks of land that used to be dedicated to railways, sidings and rail yards, mostly gone and replaced with large buildings.


Recording fidelity was much better by 1985 - sound was much richer, production values better, video was almost exclusively in color, and of course there was a lot more of it all, and in a much wider variety than in the 1950s, so there is more to choose from for later generations. Movies became more sophisticated too, both technically and artistically.


A niggle on the term "transgendered" (via http://www.glaad.org/reference/transgender), specifically the final point:

Problematic: "transgendered"

Preferred: transgender

The adjective transgender should never have an extraneous "-ed" tacked onto the end. An "-ed" suffix adds unnecessary length to the word and can cause tense confusion and grammatical errors. It also brings transgender into alignment with lesbian, gay, and bisexual. You would not say that Elton John is "gayed" or Ellen DeGeneres is "lesbianed," therefore you would not say Chaz Bono is "transgendered."


Can I ask a question? It seems surprising to me that we should compare transgender to lesbian, gay, and bisexual. Shouldn't we compare transgender to male, female, and other genders? It seems to me that being transgender is about an inward-facing personal identity, whereas lesbian, gay, and bisexual are about an outward-facing sexual preference. But these issues can be sensitive and sometimes not intuitive, so I'm never sure what to do except to try and learn the right thing.


Thank you for the education. Fixed in comment.


The difference is a person is born gay, they don't need an operation to become gay. The Ed reflects this process. Hence you cannot be gayed, but you can become educated, indoctrinated, experienced, and transgendered.


Transgender does not mean having a sex change operation. Statistics say that somewhere between 66 and 90 percent of transgender people do not get sex change operations[1][2], and 40% have any no medical treatments at all (hormonal, surgical... anything) related to transitioning their gender[1].

[1] http://www.thetaskforce.org/static_html/downloads/reports/re...

[2] http://www.nytimes.com/2011/03/23/nyregion/23gender.html


Got to big maze and... "Connection Lost"


Ah, the old "lie of omission" gambit.


According to http://www.tomsguide.com/us/chip-and-pin-credit-cards,review... "U.S. retailers have until Oct. 1, 2015, to install chip-and-PIN compatible card readers at stores. After that date, merchants will be held liable for any fraudulent charges resulting from misuse of magnetic-stripe cards."


Merchants are already de facto liable for fraudulent charges.


I still have an iPhone 4 and don't have Safari crashing issues.


The trick to speed control is to set the speed limit 10-20 mph slower than you expect traffic to actually go, and never or infrequently enforce it until it exceeds +20mph. This is the M.O. for many places already.


Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: