Thursday, January 22, 2009

Marshmallows as a personality test

The almost always interesting psych professor from Stanford, Philip Zimbardo, has a new book about how we experience time. This is indeed the guy you read about in your intro to psych class about the "Stanford Prison Experiment".

One of the things he talks about is another famous experiment tempting four year olds with marshmallows. You can see the brief video (and pitch for his new book) here. They retested the children 14 years later, and the kids that could delay marshmallow gratification were ... well I won't spoil it for you.

If you want more, there is video of his November 2008 presentation to the Commonwealth Club. Too long, but he also talks about how anti-drug programs like DARE don't work (and about addiction in general), risky driving, Monterey Bay sardines, and predictors of whether patients will complete their physical therapy.

At about 50:30 in the Commonwealth Club talk, Zimbardo talks about Fresno State prof's Bob Levine's book A Geography of Time. It is a different approach to time, and has some interesting culture-clash stories.

Bonus: More famous psych experiments you may have heard of: Elliot Aronson's "jigsaw classroom" technique for cooperative learning, and Stanley Milgram's shocking experiment.

Exercise for the reader: Design an experiment using marshmallows to screen applicants for software engineering jobs :)

Wednesday, January 21, 2009

Ah, multicast video

Was the inauguration a good case for multicast video? This summary from Merit Network said
Multicast connections, where multiple users are served by a single outgoing connection, provided Merit's members who are mulitcast-enabled with a more consistent rate of traffic, while campus networks that served unicast streaming traffic, with one stream for every single connection, experienced an exponential spike in traffic.

"Our Members who took advantage of multicast to view the Inauguration not only got great service, but didn't contribute to the traffic surge," Welch said.

"The impact of using multicast was significantly less for those Merit Members without a doubt," added Bob Stovall, vice president of network operations and engineering.
The graphs of traffic are interesting.

So is multicast the answer? Maybe not. Check this out: sending traffic from New Zealand to San Francisco and back to New Zealand because of cost.

Note to self: here is the online form to request CENIC to turn on multicast. Here's some of the 'channels" that are out there now.

Monday, January 19, 2009

Cockpit dynamics, undersea cable, national security

Another great post by Bob Sutton, this time about the recent airliner crash in the Hudson River and how no one died. He cites work by one of his mentors about the NASA cockpit studies I've talked about. Some highly excerpted quotes:
Richard [Hackman] is the world's expert on group effectiveness and I am lucky to count him as one of my mentors ... Some people in the groups area are more well-known in business circles, but Richard is the best, especially if you care about evidence rather than faith-based practices.

One of Richard's biggest research projects was on the dynamics of airline cockpit crews, and he devoted many days -- for over a year of his life -- sitting in the jump seat of the cockpit and observing and coding the dynamics if the dyad or triad. One of the main lessons that came from this --and related -- research is that the less time that a crew has been together, the more group dynamics problems they have and the more mistakes they make.
Bottom line: tired crews who've worked together make fewer mistakes than fresh crews who are new to each other.
As Hackman writes, he sometimes has the impulse, when boarding a commercial flight, to stick his head in the cockpit, and ask the crew if it is there first flight together -- the odds against an incident are very low even on first flights, but Hackman points out that your risk as a passenger would be far lower if you avoided first flights together.

Funny/sad bonus topic: AT&T is proposing bringing more transpacific fiber connections from Hawaii and Asia into their facility near Montana de Oro on the central California coast. You can see the environmental study (adverse effects are minimal since the conduit and facilities is already in place). But the funny/sad comments are here. In case they go away, here's the funny one -- "Is this necessary? I talk to Hawaii all day long using VOIP. We really don't need anymore fiber optic cables. Landlines are on the way out" -- and the both at once uninformed (there are already fiber optic cables coming onshore at that location) and corrective comment -- "It certainly doesn't sound good for Montana De Oro and I'm also against that location choice. That being said, fiber optic cables are the necessary 'backbone' that connect wireless and wired voice and data everywhere - how do you think your VOIP gets to Hawaii?"

There's a nice detailed list of Pacific undersea cables here. More information about the Montana de Oro and the Grover Beach cable landings on this guy's blog.

Another bonus topic, this one pretty sad, and something the new administration will need to pay attention to: the security situation in Mexico (see AP story, Reuters column, and WSJ column).

Wednesday, January 14, 2009

Good enough software

Steve's post about zen and software for some reason reminded me of James Bach, the famous software consultant. Note that the name of Bach's website is He's known for the idea of good enough testing and software, but "good enough" in a good way -- see the Herb Simon idea of satisficing.

Before you know it, Steve will be posting about the application of Carse's Finite and Infinite Games to software development -- but wait, Peter Denning and Alistair Cockburn beat him to it :)

Bonus Alan Kay quote (from when he was leaving Atari): "'I guess the tree of research must from time to time be refreshed with the blood of bean counters.''

Monday, January 12, 2009

Rebooting and security

This week in San Jose the National Science Foundation is bringing together academics and industrialists at the Rebooting Computing: The magic and beauty of computer science symposium. The list of participants is amazing. Here is an example, but you'll have to sign up (free) on the site for an account, and then login before you can see this -- a forum thread where Alan Kay, Gene Spafford, Grady Booch, and Dennis Frailey talk about open source software, security, and trustworthyness. After you login in, here are links to:
  • Frailey on defense contractors and FOSS
  • Spafford's response, talking about security
  • Alan Kay (and if you scroll down, Booch and Vint Cerf) respond.
One of the better signal to noise ratio discussions I've read in a while.

I've also posted two comments to Steve's post about security. No guarantee about my signal to noise ratio :)

Saturday, January 10, 2009

Software engineering, inauguratorial regalia

Was JFK the last president to wear a top hat at inauguration? If you are a little late putting together your wardrobe for the 20th, I recommend this or if you are feeling more patriotic, this.

The IEEE Computer Society is commemorating 25 years of Software magazine. Some of articles are available to nonmembers and some are interesting (feel free to draw your own Venn diagram):
  • C's Brian Kernighan on sometimes the old ways are best (he complains about linux breaking wc)
  • Niklaus Wirth's (father of Pascal) brief history of software engineering
  • and the "top 35" articles published in Software so far. I think I have just found the syllabus for a intro software engineering graduate class :)
Bonus: A free 2009 downloadable calendar from Dr. Dobbs. Believe me, if you are a software developer, you need it for your cubicle.

Bonus bonus: Technology Review created an oral history of space tourism by interviewing five of the six Soyuz passengers to the space station and interleaving their responses. Just like with commercial airlines, Anousheh Ansari's luggage was lost and she ended up with men's shaving cream and cologne instead of her stuff.

Doesn't it seem like microgravity would be good for back pain? But astronauts frequently complain of lower back pain maybe because the spine decompresses.

Thursday, January 08, 2009

Methodist Ministers and other topics

Assorted things:
  • Bob Sutton uses a study of Methodist ministers to draw conclusions about leadership.
  • The Dirty Jobs host talks about peripeteia, anagnorisis, and finding value in your work.
  • One of my neighbors has a stinky fireplace, I hope they are consulting this.
  • Kathy Sierra from the 2008 O'Reilly emerging technologies conference mentions the Alan Kay video I've talked about previously.
  • And Kent Beck from the 2008 O'Reilly Rails conference retrospects about the last 20 years of his work in extreme programming, Smalltalk (there's Alan Kay again), design patterns, and the influence of physical architecture.
  • Guy Kawasaki at the November 2008 Commonwealth Club on "No Bull Shiitake". He talks about Steve Jobs, Woz, Obama, the best part of working at Apple in the early days, funding young entrepreneurs.