Do You See a Pattern?

http://www.slate.com/id/2237109/

by edw519 • 17 years ago

84 30 17 years ago

ojbyrne• 17 years ago

I really enjoyed "A Pattern Language, " even though it wasn't relevant to my life other than its influence on the Design Patterns people.

cpr• 17 years ago

Gosh, if you've ever even thought about building your own home, APL is invaluable.

I think the book should be required reading for everyone in high school, and the language acquired with enough fluency that you could "speak" your own house into being.

Fantastic stuff.

His magnum opus (4 volumes) The Nature of Order is rather daunting, by comparison, but probably quite important. I've got it, but haven't had time to really dig in...

nfnaaron• 17 years ago

I never would have been aware of APL without being in software, and I'd never read anything on architecture or building before that. It was one of the most interesting books I've ever read, and it made a lot of sense. Yes, if I were building my own home, I would definitely re-read it cover to cover.

RyanMcGreal2• 17 years ago

> The theories that are taught in architecture schools today are of a different sort, and in the belief that the field of architecture should be grounded in intellectual speculation, rather than pragmatic observation, students are more likely to be assigned French post-structuralist texts than A Pattern Language. Which is a shame.

It's just astonishing, decades after Jane Jacobs' The Death and Life of Great American Cities, that pragmatic observation is still on the margins. What, a half-century of abject failure in building and neighbourhood design isn't enough to demonstrate the fundamental wrongheadedness of dogmatic, a priori design principles?

david927• 17 years ago

It's a political problem; the democratization of principles. Sometimes what everyone likes isn't what's best for everyone.

It's interesting to note that Kent Beck, who brought Alexander to the attention of Erich Gamma, did so with the intention that it could be used for interface design (if I remember right).

donaldc• 17 years ago

Such is the power of institutional, organizational, and societal inertia.

notaddicted• 17 years ago

Would you advocate the same change in Computer Science education? Enough intellectual speculation: Java schools for all.

To make progress, you have to try something new.

RyanMcGreal2• 17 years ago

> Enough intellectual speculation: Java schools for all.

A monolithic, one-size-fits-all approach is hardly consistent with an evidence-based approach that considers the needs of individual cases.

Skeuomorph5• 17 years ago

There's an unforeseen problem with well patterned livable houses--they often don't sell well to conventional real estate buyers. If you need to relocate or access equity quickly, you're stuck with an remarkable property but buyers looking for McMansions. Buyers that do drop by out of curiosity comment with gems such as, "We could never live here, the master bath needs to be as big as the bedroom."

Seems as though the only way to move such a property is to offer it so far below market that people become willing to compromise and consider the more livable home instead of yet another generic "keeping up with the Joneses" colonial:

http://images.google.com/images?q=colonial%20house (exteriors)

It's galling, because such easily substitutable boxes have little intrinsic value beyond market. The cost to build a colonial is quite low compared to the cost of building, for example, "... varying ceiling heights between large and small rooms to create different degrees of intimacy."

One eye-line in my home reveals 8 ceiling heights, correlating to the intimacy of the spaces below them. It's extraordinary, but generic buyers want five bedrooms with full baths so what should be livable spaces end up the size of closets.

What's worse, when buyers today see the prices of McMansions plunging, they expect to see the price of a well patterned home (that may have cost 5x - 8x as much to build) also plunge, despite the more complex home already having a replacement cost higher than its asking price, price of materials and labor having risen.

(PS. Serious inquiries welcome for a well patterned home a commuter train ride from Manhattan.)

kurtosis• 17 years ago

I agree with you nearly 100%, but why do you think the market fails to recognize the higher value of such a home?

anamax2• 17 years ago

> Seems as though the only way to move such a property is to offer it so far below market

You don't understand the term "market".

> What's worse, when buyers today see the prices of McMansions plunging, they expect to see the price of a well patterned home (that may have cost 5x - 8x as much to build) also plunge, despite the more complex home already having a replacement cost higher than its asking price, price of materials and labor having risen.

There's a cost associated with being "goofy". If that cost exceeds the benefits, perhaps you should reconsider your decisions.

ExcilSploft• 17 years ago

sans buying the book (which I now would like to), can you provide an example of a well patterned home as you did for Colonial/McMansions?

rjurney• 17 years ago

Me and two other people built a house without blueprints. We just thought out the internals as we built it, debating the layouts by pretending to take actions on the bare floors. We got most things right, and a couple things wrong (doors in bad spots - why oh why not 2 feet right?). There are very few 90 degree angles (by design). Its an amazing house, very livable, and it suffers the same problem as yours.

martian• 17 years ago

I recommend Notes on the Synthesis of Form to everyone. For one, it's a delightful read (Alexander is a master linguist), but it's also tremendously helpful for anyone thinking about building a large system, especially software. The main takeaway for me was thinking in terms of systems of constraints. When designing a vacuum, constraints might be: needs to be reasonably priced, needs to be profitable, needs to be easy to carry, needs to be durable, needs to be powerful enough, etc. Adjusting one constraint usually affects another, so it's easy to see how complicated this can get.

Alexander formalizes the process, and demonstrates mathematics to solve the constraint problem. This assumes, of course, that you can articulate and quantify the constraints in the first place.

Web developers are lucky: unlike architects or product designers they can quickly test their assumptions (A/B tests) and move the system to a better state.

bengebre• 17 years ago

I second the recommendation. A Pattern Language outlined a lot of the actual patterns (which was neat), but Notes on the Synthesis of Form did a better job of illuminating a broad conceptual foundation for designing systems. I think I need to re-read it now that you mention it.

mbrubeck3• 17 years ago

I'd also recommend The Timeless Way of Building, which provides the overview of Alexander's method (of which A Pattern Language is just one piece). I once wrote a review of The Timeless Way from a programmer's perspective: http://advogato.org/person/mbrubeck/diary/106.html

(The Oregon Experiment, which I mention in that link, might also be of interest to programmers inside huge organizations.)

hugothefrog• 17 years ago

Christopher Alexander also wrote the introduction to Patterns of Software (http://www.amazon.com/Patterns-Software-Tales-Community/dp/0...).

It's a great book, and I've unfortunately only had the time to read a few of the essays.

It's available online as a PDF from http://www.dreamsongs.org/Files/PatternsOfSoftware.pdf.

gojomo• 17 years ago

I have great respect for Alexander's ideas, but when he's mentioned, I can't resist sharing this cheeky presentation from one of his Berkeley hills neighbors about the "patterns" evident from the exterior of Alexander's house:

http://metatime.blogspot.com/2005/04/patterns-good-and-bad.h...

(I don't believe the presentation was prepared with the intent to belittle Alexander's work, and I don't share it in a negative spirit -- but rather as a sort of "only human"/"cobbler's children" observation about the contrasts between a professional oeuvre and personal life.)

UncleOxidant• 17 years ago

Wow...that's... really... his house? What a mess. I can't see how a renowned architect would want to live in such a ramshackle way. Seems like if you're an architect you need to have your house be a showcase of your work - an example that clients could look to get an idea of some of your better work.

mbrubeck3• 17 years ago

That's exactly how I'd expect Christopher Alexander's house to look. The buildings he praises in his work include few cathedrals, and many folk houses in traditional villages around the world. He wrote that the clean, sterile lines of Modernism were designed to enhance the egos of architects, not the lives of their clients. The houses he built in Mexico (The Production of Houses) were made cheaply, using community labor, from handmade local materials.

He also believes that a building is never finished, and should be continually modified by its inhabitants. (Stewart Brand's How Buildings Learn is another book that should be read by both programmers and architects.)

tome• 17 years ago

The documentary version of How Buildings Learn is also available:

http://video.google.co.uk/videoplay?docid=863955592548621085...

praptak• 17 years ago

"Design Patterns aren't" is required reading for anyone interested in both Alexander's work and software design patterns: http://perl.plover.com/yak/design/

Same goes for Alexander's speech at The 1996 ACM Conference: http://www.patternlanguage.com/archive/ieee/ieeetext.htm

kqr2• 17 years ago

Here's a summary of his influential book A Pattern Language:

http://downlode.org/Etext/Patterns/

Do You See a Pattern?

Related Stories