City Layout Design: A programmer's perspective
thedissectionroom.com
thedissectionroom.com
The argument is roughly the same as the one against waterfall development in software engineering which is that it's impossible to know until actual implementation & use what is successful or not.
Instead, successful cities are evolved and mixed use, with a mix of new and old buildings, strong social fabric and buildings built to be adaptable to changing needs.
See: How Buildings Learn, The Death and Life of Great American Cities and A Pattern Language for more context on this.
The Nature of Order series generalize even more, so I'd read them after that. They rather independent from the TWoB and APL but I believe the way from the concrete examples (APL) to the generalizations is a better one. Here it's rather a matter of preference than one of forward-dependency.
There's place for organical growth, and there's place for upfront design, both in programming and in urban development.
The best cities layout I've seen was consequences of organical deveopment within some rules - see most old european cities. Building rules specified how broad should streets be, how high the building can be in each region, which functions can't be performed in which part of the city, which way should new streets be built. There were reasons for these rules.
In my country after 50 years of communism (and hideus commie blocks, etc) we went throught 10 years of "free for all" development of cities.
Believe me - I hate totalitarism, but in these first 10 years of freedom cities got much worse - wallmarts in the city center, ugly cheap postmodern buildings near 500 year beutiful houses, commercials hiding half of the city besides them. Now, since +- 10 years there is some regulations coming back, and I'm happy with that.
I've felt that organically grown cities, of which most major European cities are good examples, are very sub-optimal for because of the haphazard and random way the cities are laid out. It many such cities, the old, medieval core became so bogged down that cars aren't even allowed and it's basically a pedestrian heaven.
Navigation is hard in these cities, traffic is usually a mess without modern development to route around the "organicness" of the city centers, as a result certain types of goods and services...namely anything to do with lots of manufacturing or bulk goods, ends up moving away from the city core. And while it means that the cores tends to become very pedestrian friendly, they really just sort of end up as extended outdoor shopping malls with some housing for the shop workers above.
It's interesting in some of the older Asian cities which have very old histories, say Seoul, the the haphazard, medieval-style organic planning has given way to a large effort to put the major thoroughfares on a grid and as old areas are redeveloped, to do so with a heavily planned development policy in mind. The city was simply not able to handle the growth and population demands the organic system created.
There's a reason Paris' business center is not in Paris' geographic center, and central London has restricted driving zones, and most of Barcelona's population enjoys the 19th century planned developed residential areas, while the older parts are being turned into tourist hotels, malls, restaurants and more shopping areas.
Pedestrian streets in the center, industry forced out of the city, services concentrating in the center, cars not viable for in-city transport, well developed public transport.
Also, I don't think medieval and older cities were not planed. They were just designed without cars and milions of people in mind. With cars I think it's actualy good - cars and cities don't mix well.
BTW: I know whole city designed by one architect as "ideal town" in 1580, and built according to these plans. It's quite nice to live - http://en.wikipedia.org/wiki/Zamo%C5%9B%C4%87
It's interesting how many old cities were started as a heavily planned Roman garrison towns, straight boulevards, right angles everywhere, etc. Then those cores turned into the medieval rats nests we're familiar with today, then usually outside of that there was a building boom in the 19th century and a return to some kind of order.
But the advantage of the unplanned medieval layout is that the streets are usually an attempt at a shortest path between two interesting points.
A better configuration would be three spiral shapes, giving the route defined start- and end-points to allow recovery time for delays.
[edit] I should mention that overall, I quite like the design. It looks to me like you've arranged the space at a livable scale and with careful attention to the work/live/play cycle of its inhabitants.
So transport routes, parkings have to go throught the city, dividing it, using up space, making everybody go further and further to do anything, and isolating people in the center of the city from the nature.
So he imagined cities that are like lanes - growing from the center in only 2 directions. Everybody will have nature near, between such urbanised lanes will be villages, roads, etc - so transport will go throught wilderness, making it cheaper, and safer. These lanes should be continious, and areas of different functions should be placed repeatedly, in small distances, so nobody will have to go too far from home to servicing areas.
Of course - it's a little utopian idea, but still, I like it.
Some more info about Oskar Hansen, and his architecture:
http://tnn.pl/Oskar_Hansen,2969.html
One of the districts he designed (he had to make compromises, so it's not really LCF, but still, it's nice district for its time). http://tnn.pl/uploaded/zdjecia/200806191346500.hansen_041.jp...
Every 4 or 5 miles the scene will basically repeat. The same set of stores, in virtually the same order, arranged in their various strip malls in virtually the same way. It felt like you were driving in a Zoetrope.
Go a mile or two off that and you're in the Gulf of Mexico or in a swamp if you go the other way.
It was an interesting city plan, but it felt miserable to me (other than the close proximity of nature).
http://en.wikipedia.org/wiki/File:Inner-canberra_01MJC.png
It would also be worth checking out New Urbanism, which has very similar goals for the designs of cities.
Vice: I think most people’s natural inclination after watching your video would be awe, followed by fear.
Vincent: I am in agreement. That was part of the intent of the video, for those that know, play, and love the game.
Introversion, an indie game company in the UK who created hacker favorites like Uplink has unfortunately shelved their game project played in a fully procedurally generated city, from the major neighbourhoods down to the individual floors of the buildings. You could get a demo of the city generator from the recent Humble Bundle. There are still some fascinating videos to find on their blog [3] or on youtube [4]
[1] http://youtu.be/NTJQTc-TqpU [2] http://www.vice.com/read/the-totalitarian-buddhist-who-beat-... [3] http://www.introversion.co.uk/subversion/ [4] http://youtu.be/J30i0gABfS8
[edit]: better link for the Subversion video
http://www.newworldeconomics.com/archives/2009/111509.html
Another worthwhile article is "Let's Take a Trip to Suburban Hell."
And he blasts carfree's "green space", but then lists a whole bunch of things which the green space could be (yards, parks, sports fields). I can't help but feel that he's deliberately missing the point.
There's also a fairly detailed discussion on carfree.com as far as "how to get there from here": http://carfree.com/conv_lyon.html so it's hardly billed as a 'build it from scratch' ethos either.
Public transit will lift the limit somewhat: people are willing to spend only so much time commuting or going to/from hobbies but with good public transit people will frequent other neighbourhoods on a regular basis. Public transit will be cheap as long as densities are high enough and that's why big cities are dense and have public transit. Large car-based "cities" aren't actually cities, they're just glorified urban office parks with people living in the suburbs: public transit will never work there efficiently.
The reason why you can't design a city is that a meaningful city can only grow organically based on whatever people happen to make of it. Commerce will surface when the place and time are right. Predesigning a neighbourhood will result in either a lot of dead space or a preformed area that nobody will consider their own.
This is false, though. The method of growing doesn't affect the feasibility of public transport: the density of the areas in question does.
And no matter how the areas grew, an economically growing public transport network will naturally connect the most dense parts of a city together. That is, provided there are dense areas. Think London's 100+ year history of the Underground where underground railway companies, as soon as funding was secured, built little stretches that made the most economic sense.
Also I believe the dead space problem will take care of itself as demand for space increases.
It doesn't if the environment doesn't lend itself to grow, change, and adapt as needed which is a big problem with designed areas. An area also needs both old and new lots and buildings to offer space for businesses that wouldn't survive with market rents. This is something nobody can design.
You're quite never sure of that. Just yesterday evening I was walking down on a street in my city (East-European capital) whose 19th-century buildings were raised to the ground around 25 years ago. There are still un-developed patches of land on that street, which, needless to say, make it look deserted and desolate even though it's only a 5-minute walk from the Prime-Minister's office building.
To preserve the grandeur of the National Mall, the White House, the Capitol, and various other key locations, the entire city is subject to strict building height limits. This limitation was placed in effect just prior to the 20th century when government officials realized that structural steel "skyscrapers" could overwhelm the city. In 1899, Congress enacted a height limit for the District prohibiting private buildings from rising more than 130 feet. Contrary to popular belief, no law has ever restricted buildings to the height of the United States Capitol or Washington Monument.[20][21] A revised height law in 1910 did away with that fixed maximum. The newer legislation, still in effect today, states that no new building may be more than 20 feet taller than the width of the street in front of it.[22] The current law is codified as D.C. CODE ANN. § 6-601.05. Thus, Washington has a relatively modest skyline in comparison to the majority of American cities. However the District is ringed by high-rise buildings in many nearby suburbs like Arlington, Silver Spring, and Bethesda.
Addendum: I realized on the way home that this comment was rather too content-free for HN. So...
More specifically, I was curious if the book you mention has some narrative or historical details on why the grand Epcot vision fell apart/was stymied.
It has similar flaws, as well. It should be considered descriptive, not prescriptive--and Christopher Alexander is not nearly as careful around that distinction as the GoF.
http://www.patternlanguage.com/archives/alexander1.htm
For some commentary on that piece, see:
http://contraterrene.com/wiki/index.php/Commentary_on_%22A_C...
tl,dr: designing cities as hierarchical trees, while perhaps efficient in the sense of transportation links, is a bad idea from the standpoint of making a cohesive city.
I have structured my city into smaller circular satellite cities (C1 to C6) which in turn are placed within a circle *with the airport at the center.*
Tokyo, too, has multiple sub centres which are mostly self-sufficient. They are connected by a train running around in a circle and have central hubs where trains depart into the suburbs. Stations are largely accommodated with large bicycle parking lots to facilitate train usage. However, even though Tokyo is one magnitude bigger than NYC, the transportation fees are fairly high, too. Also, the Japanese solved the parking space problem, by making it a requirement to have a parking space in order to be able to buy a car.
The parking space problem is only solved because everyone is using the train to go to work everyday, mostly. If this were to change, they would face major parking space issues, as the innards of the city are not designed to park a great amount of car. The parking space you mentioned is usually the one you reserve in your building when you get a car, but that does not mean there is space available at your destination.
It is quite inefficient to travel in Tokyo altogether. While not too expensive, it takes a lot of time and in the train you cannot seat most of the time because it is overcrowded. I would not pick Tokyo as an example of a city where things work well - it just works, but it's already over-capacity.