codemate

The amateur scientist (that's us)

Fri, 20 Nov 2009 02:41:00 -0800

Many people buy a car (probably their single biggest discretionary purchase) based on slamming a door, kicking a tire and judging the handshake of a salesperson.

We choose a surgeon based on the carpeting in his office and a politician by his hair cut.

During the first week of swine flu vaccines in New York, most parents (more than half!) chose to keep their kids out of the program.

Interviewed parents said things like, "I'm not sure it's safe," and "I wanted to see if it affected other kids..."

No mention of longitudinal studies or long-term side effects. No science at all, really, just rumors and hunches and gut instincts.

This gut-instinct approach served people well for hundreds of thousands of years, but it's pretty clear that it doesn't work in a complex world. Eating salmon at a wedding feels 'safe' because we always have, but of course any professional scientist will tell you that farmed salmon is an ecological disaster. You can't see the problem, so you ignore it.

Audiophiles spend thousands of dollars rewiring the electrical lines in their house with .99999% pure copper, ignoring the fact that the power from the street is in the same old cables. Adding decimal points to our irrationality doesn't change much.

The problem with being an amateur scientist is precisely the reason that marketers relish the opportunity to sell to us, the amateurs: we make stupid decisions, easily manipulated by those who might choose to do the manipulation (on their behalf or on ours).

The news here is not that people are irrational, giving too much credence to the dramatic and the local and the short-term (that's not news), but that people have added a veneer of scientific rationality to their irrational decisions. Armed with Zagats or internet data or some rumor off Snopes, we act as though now we're supremely rational choicemakers.

This is one of the problems with breast cancer screening. It appears to give information, really good information, but in practice, it doesn't. Since the information is vivid, we give it too much credence.

The challenge for people trying to market vaccines or highlight long-term side effects of various consumer choices is that it's much easier to spread a story about exploding cars or hair falling out than it is to spread a story of 'nothing bad happens' or 'no one got the swine flu and died' or 'three years from now, this section of ocean will be dead.' We prefer the vivid anecdote to the dry and statistically useful fact, which in a complex world, is to our detriment.

PS if I was marketing the swine flu vaccine, I'd name it after a kid who died last season and put her picture on the release form. Alas, teaching amateurs like us to be real scientists is going to take a while.

Vim Recipes: Opening the File Name Beneath the Cursor

Tue, 17 Nov 2009 17:31:03 -0800

What's the best / most interesting piece of code you've ever seen?

Sun, 15 Nov 2009 23:06:36 -0800

Teaching Kids Programming

Tue, 29 Sep 2009 23:41:09 -0700

http://scratch.mit.edu/

Coordinate conversions made easy

Tue, 22 Sep 2009 10:17:46 -0700

Teenager builds robot to solve his Rubik's cube

Tue, 15 Sep 2009 05:04:33 -0700

The hierarchy of success

Mon, 14 Sep 2009 03:32:00 -0700

I think it looks like this:

Attitude
Approach
Goals
Strategy
Tactics
Execution

We spend all our time on execution. Use this word instead of that one. This web host. That color. This material or that frequency of mailing.

Big news: No one ever succeeded because of execution tactics learned from a Dummies book.

Tactics tell you what to execute. They're important, but dwarfed by strategy. Strategy determines which tactics might work.

But what's the point of a strategy if your goals aren't clear, or contradict?

Which leads the first two, the two we almost never hear about.

Approach determines how you look at the project (or your career). Do you read a lot of books? Ask a lot of questions? Use science and testing or go with your hunches? Are you imperious? A lifehacker? When was the last time you admitted an error and made a dramatic course correction? Most everyone has a style, and if you pick the wrong one, then all the strategy, tactics and execution in the world won't work nearly as well.

As far as I'm concerned, the most important of all, the top of the hierarchy is attitude. Why are you doing this at all? What's your bias in dealing with people and problems?

Some more questions:

How do you deal with failure?
When will you quit?
How do you treat competitors?
What personality are you looking for in the people you hire?
What's it like to work for you? Why? Is that a deliberate choice?
What sort of decisions do you make when no one is looking?

Sure, you can start at the bottom by focusing on execution and credentials. Reading a typical blog (or going to a typical school for 16 years), it seems like that's what you're supposed to do. What a waste.

Isn't it odd that these six questions are so important and yet we almost never talk or write about them?

If the top of the hierarchy is messed up, no amount of brilliant tactics or execution is going to help you at all.

An Industrial-Strength Audio Search Algorithm

Sat, 05 Sep 2009 22:18:05 -0700

How Shazam works.

Objective-C Programmers: Whats so great about Objective-C?

Fri, 04 Sep 2009 19:02:09 -0700

Reddit thread.

Mac OS X Automation

Wed, 02 Sep 2009 21:40:29 -0700

Bunch of services for MacOS X

Mac OS X 10.6 Snow Leopard: Review

Tue, 01 Sep 2009 11:41:21 -0700

Michael C. Feathers: Seven Blind Alleys in Software Design

Tue, 25 Aug 2009 10:04:01 -0700

BeanKeeper ORM framework for Java

Tue, 25 Aug 2009 16:40:39 -0700

How to Hack the Microsoft Job Interview

Tue, 25 Aug 2009 16:27:44 -0700

http://halcyon.usc.edu/~kiran/msqs.html

How to Debug Bash Scripts

Mon, 24 Aug 2009 21:10:36 -0700

Debugging tips for Objective-C programming

Thu, 20 Aug 2009 21:46:59 -0700

Geonames Reverse Geocoding

Thu, 20 Aug 2009 12:53:25 -0700

http://ws.geonames.org/findNearbyWikipediaRSS?lat=49.256154&lng=-123.083267&radius=50&maxRows=100

iphonearkit - iPhone Augmented Reality Kit

Thu, 06 Aug 2009 12:51:43 -0700

So you've just been hired by an IT department... ‎(Software Engineering Tips)‎

Thu, 06 Aug 2009 12:31:09 -0700

fun to read and mostly true.

Beta testing on iPhone 2.0

Wed, 05 Aug 2009 22:54:55 -0700

Why? Language Archaeology ... and Metaprogramming

Wed, 17 Jun 2009 05:01:40 -0700

Grant Rettke: A Card Stack Handheld UI Approach

Sat, 06 Jun 2009 18:39:49 -0700

Guy Steele: The Future is Parallel, What's a Programmer to Do? [PDF]

Wed, 03 Jun 2009 21:59:41 -0700

Are the new JavaScript engines going to completely change the game?

Tue, 02 Jun 2009 21:13:34 -0700

E-Book: A Functional Pattern System for OO Design (How FP can advance OO)

Tue, 26 May 2009 10:48:50 -0700

Reverse engineering protection: code so awesome, you can't even _look_ at it without crashing.

Tue, 26 May 2009 07:51:25 -0700

A Coder's Guide to Coffee

Fri, 22 May 2009 01:05:36 -0700

Fun with Type Functions : Oleg Kiselyov, Ken Shan, and Simon Peyton Jones' new paper: Feedback wanted!

Sat, 16 May 2009 11:52:29 -0700

Author of "Modern C++ Design": STL iterators must die

Thu, 07 May 2009 13:58:37 -0700

The Good, the Bad, and the Ugly: The Unix Legacy (by Rob Pike)

Thu, 07 May 2009 04:04:41 -0700

Fun names. A Haskell function called isn't

Wed, 06 May 2009 07:21:34 -0700

A comparison of C++ "concepts" and Haskell type classes :: PDF

Sun, 12 Apr 2009 09:47:29 -0700

Beautiful Horizontally Centered Menus/Tabs/List. No CSS hacks. Full cross-browser.

Fri, 10 Apr 2009 04:09:18 -0700

April Fools Roundup so far

Wed, 01 Apr 2009 12:10:50 -0700

Interview with the man who built Google's V8 engine from scratch

Sat, 28 Mar 2009 12:45:39 -0700

The Pragmatic Programmer Quick Reference Guide

Tue, 03 Mar 2009 10:35:27 -0800

Color Scheme Designer 3

Wed, 18 Feb 2009 11:54:02 -0800

Best Open Source Resources for Web Developers | WebAppers

Wed, 18 Feb 2009 11:53:07 -0800

10 Useful Web Application Interface Techniques

Wed, 18 Feb 2009 11:48:48 -0800

How To Convert From a VMware Image To Virtualbox Image

Sun, 28 Dec 2008 09:09:00 -0800

OK and Cancel Buttons: What's the Right Order?

Thu, 18 Dec 2008 05:07:19 -0800

Hibernate: Truly Understanding the Second-Level and Query Caches

Tue, 16 Dec 2008 10:29:38 -0800

Hibernate: Improving Performance (Caching)

Tue, 16 Dec 2008 10:28:08 -0800

improving hibernate performance with cache

Being Joe the Software Plumber | Javalobby

Sun, 07 Dec 2008 07:29:42 -0800

Enterprise Integration Patterns

Wed, 19 Nov 2008 04:15:56 -0800

Classycle - Analysing Tools for Java Class and Package Dependencies

Sun, 05 Oct 2008 07:04:25 -0700

Classycle's Analyser analyses the static class and package dependencies in Java applications or libraries. It is especially helpful for finding cyclic dependencies between classes or packages. Classycle is similar to JDepend which does also a dependency analysis but only on the package level. Classycle's Dependency Checker searchs for unwanted class dependencies described in a dependency definition file. Dependency checking helps to monitor whether certain architectural constrains (e.g. in a layered architecture) are fulfilled or not.

JIP - Java Interactive Profiler

Thu, 02 Oct 2008 11:53:05 -0700

Producing Open Source Software

Wed, 01 Oct 2008 05:49:38 -0700

How to Run a Successful Free Software Project

Useful Eclipse Plug-ins - Codehaus

Mon, 29 Sep 2008 06:37:19 -0700

Joda Time - Java date and time API

Thu, 25 Sep 2008 01:25:40 -0700

Joda-Time has been created to radically change date and time handling in Java. The JDK classes Date and Calendar are very badly designed, have had numerous bugs and have odd performance effects