Random ramblings from the Garrity family
NASA parked the Spirit rover this week. It had gotten stuck in the sand 10 months ago, they haven’t been able to come up with any way to get it unstuck, and winter’s setting in. So after 6 years in a mission which was supposed to last 3 months, they’ve finally shut it down. That’s an awfully good run.
Our favorite take on this was today’s XKCD comic. Aww….
Ed Thigpen died yesterday. He was the drummer on so many of those great Oscar Peterson records. Together with Ray Brown, they defined what the Jazz trio should sound like.
Tonight at dinner, I played their record with Ben Webster, but I can’t find that online, so go check these out instead:
I recently read Wynton Marsalis’ book Moving to Higher Ground: How Jazz Can Change You Life. I recommend it if you want to know what Jazz is about. The amazing thing about groups like Oscar Peterson’s trio is the way the performers are listening to each other and can respond and change directions instantly. The skill and attention which are required to be able to do that just staggers me.
There was an interesting paper recently where they placed Jazz musicians in an MRI while they improvised and measured what parts of their brains were active when they did this. This is your brain on jazz.
We went to see Gatz at the ART the other night.
Gatz is a play by the Elevator Repair Service which is based on the novel The Great Gatsby. It’s a very interesting production. It takes place in a fairly generic office. One of the characters is waiting for his computer to be repaired, and he picks up a book and starts reading. It’s a copy of The Great Gatsby. As he reads, the other actors slowly start becoming characters from the book. He reads the entire book. The play is something like 7 hours long. It turns out to be a very compelling presentation. It has a lot of the visual interest of the theatre without losing the beauty of Fitzgerald’s writing.
J. D. Salinger died yesterday. The Catcher in the Rye and The Great Gatsby would make most people’s short list for The Great American Novel. Nick and Holden, together with Huck Finn, pretty much set the mark for what we think of as the voice of the young American male.
They’re also novels which have a reputation for changing people’s lives. I can’t say that either of them had that effect on me. I think that these are books that resonate for people if they’re read at a very particular time in their lives. That time when you’re just starting to become cynical about what’s possible in life.
Both authors found it hard to deal with the aftermath of creating the very thing that they had so hoped to create. As Fitzgerald say so poignantly:
There are no second acts in American lives.
The Utne Reader has an interesting interview with Jonathan Haidt about politics. If you’re not familiar with Jonathan Haidt, he’s a Professor of Social Psychology at the University of Virginia. He’s done a lot of interesting work on the science behind morality. This interview is an interesting introduction to his attempts to apply this research to politics.
The basic idea behind his theories about morality is that people have intuitive ethics in 5 different areas.
And that the differences in morality between cultures or groups can be explained by different weighting of these 5 factors. If you go to his site moralfoundations.org, you can take a quiz which will measure your weighting of these factors.
He’s recently been applying this research to understanding the differences between liberals and conservatives. His premise is that our political stance is driven by these intuitive ethics, but that we’re mostly unconscious of that. He argues that the only way to get past divisions in politics is to understand the driving force between the other person’s point of view, even though they don’t know what that force is themselves.
You can also get his thoughts on these ideas from the Ted talk he gave in 2008.
I would also recommend his book The Happiness Hypothesis. It’s primarily about why we’re so bad at predicting what will or will not make us happy. But it is related to his morality studies because in both cases his research shows that an unconscious part of our mind decides what’s good or bad and then a part closer to the surface makes up a story about why. This leaves the rational part of our brain thinking that it knows the reasons for what we like and dislike when it really doesn’t have a clue.
Tim Bray wrote a great post about being invited to visit his fifth grade son’s class and “teach the internet”. He created a blog and let them all start posting. Go check it out.
Here’s another SPDE sketch. It’s a simple image warping demo that shows how to use the texture mapping feature of Processing. Click and drag over the image to distort it. Try different mouse buttons.
Source code: warp
Built with SPDE and Processing
This is similar to an image warping demo I wrote more than 20 years ago at Stellar when we launched the GS1000. I remember demoing it at one of our first trade shows and Ed Catmull stopped by the booth to talk. This was when Pixar had just gotten spun off from Lucasfilm, way before Toy Story, so he was mostly known for cool software like this image warping algorithm. He watched the demo for a while, and then said to me:
You’re just brute forcing it, aren’t you?
Yes, that’s exactly what this is doing! His algorithms are very elegant. They’re careful to not transform any pixels more than necessary and to walk through memory in an efficient order. We were able to brute force it on the Stellar because we had a ridiculous amount of memory bandwidth for the day.
Of course today the GS1000's numbers don’t seem very impressive. That’s the way it always goes. As the tide of hardware performance rises, you get to a point where it is possible to solve some interesting problem (e.g. image warping, video decompression, etc.) by carefully using elegant algorithms. Then the tide rises a little more and solving the problem becomes trivial and we all move on to the next problem.
Yesterday and today we were making some bread. We weren't making normal bread we were making no knead dough bread. This bread can "be made by a four year old". The recipe is really easy and you can check it out here at Breadtopia. The New York Times did an article on this in 2006. The bread was really great and easy to make. I will probably make a lot of of this bread from now on.
The original New York Times video doesn't work any more so here it is
There have to be a thousand recipes for Chicken Divan. I’ve been using a quick Chicken Divan recipe from a cookbook put out by the Lexington Preschool PTA (now called LexFUN) as a fundraiser. But somehow it never came out quite as I thought it should, so tonight I decided to mess with it a bit, and I like this one better. It’s still very quick, about 45 minutes start to finish if you have everything ready.
Chicken Divan
1 lb boneless chicken breasts
2 heads of broccoli, cut into florets
1/2 cup chopped onion
2 cups shredded cheese
1 can cream of mushroom soup
2/3 cup sour cream
2/3 cup mayonnaise
1/2 cup milk (approx)
3/4 cup breadcrumbs (preferably panko)
2Tbsp butter
salt & pepper
Preheat the oven to 350 degrees. While the oven is heating, steam or parboil the broccoli florets – DON’T OVERCOOK! The broccoli should definitely still be crisp or it’ll be too mushy after it’s baked. Sauté the check breasts until browned on both sides, then cut them into slices. [I guess you could short cut this part by buying pre-cooked chicken strips]. While cutting up the chicken sauté the onion until just translucent. Mix the onion with the chicken.
For the sauce mix the mushroom soup, mayonnaise, sour cream, and cheese. Add enough milk to get a thick creamy texture. Melt the butter in a small bowl, then add the breadcrumbs and mix until the breadcrumbs are all buttery.
Grease a 9x13 pan. Put the broccoli in the bottom. Layer the chicken and onions over the broccoli. Pour the sauce over the chicken. Top with the buttered breadcrumbs. Pop the whole thing into the oven for 30 minutes. Serve with egg noodles or rice.
I was actually thinking that adding some chopped red pepper along with the onion would add a nice contrasting color – maybe I’ll try that next time.
There’s been a lot of talk recently about Murdoch spreading the craziness from the op-ed section of the Wall Street Journal into the reality based sections. This is what a lot of people feared would happen when he bought it, but it really didn’t seem very likely. Doing that would be crazy, wouldn’t it? The Journal’s reputation was what he paid $5.6 billion for.
I haven’t been reading it lately, and I’ve been willing to discount articles like David Carrs’, but now Barry Ritholtz is saying the same thing. I don’t think he has an ax to grind, so perhaps it really has happened.
Oh well, I guess that completes our transition to a society where Jon Stewart is our most trusted newscaster.
William Gurstelle has a nice short article about Arduino in the latest issue of The Atlantic. He’s trying to give their audience an idea of what maker culture is all about.
One of the more hideous quirks of the family of programming languages derived from C is the fact that values of pointer based types may be NULL. This means that a simple code fragment like this:
String s = find(input,token);
if (!s.empty()) {
// do something
}
might segfault or throw an exception. You need to write something more like this:
String s = find(input,token);
if (s != null && !s.empty()) {
// do something
}
This is the root cause of the most common bugs I’ve seen in C, C++, and Java code, but a lot of developers in those languages seem to think that it’s just a fact of life that you have to accept.
A number of languages don’t have this problem. They simply don’t allow NULL values for these types. It turns out that it’s possible to throw the baby out with the bathwater when you go that direction. In the example above, find might really want to distinguish between finding an empty string and not finding any string at all. You can work around this with a second return value, but the result is pretty ugly. Something like returning NULL is actually cleaner in this case. But does seem crazy to burden everyone who uses strings with null pointer exceptions just because of this one use case.
Several recent languages have done a better job of balancing these two concerns. One noteworthy example is the Maybe type in Haskell, but there are good examples in a number of other languages.
Guy Steele just posted a very thoughtful article on his blog about this problem and how they approached it in Fortress.
Their basic approach is that the Maybe type is a mixin. This means that you can create a type that is either a String or nothing without forcing that option on everyone who uses strings.
One nice feature of his post is that they really chased down the loose ends. They realized that the Maybe type is related to a collection in many ways. But that it is a collection which can only have 0 or 1 items in it. If collections like that are important (and they are if you introduce Maybe), then that has implications for everything that deals with collections. For example, there’s a nice if/else syntax for this type of collection.
If you’re interested in programming languages and the ins and outs of designing datatypes, you should go read it.
Do you remember our earlier post about Gever Tulley? Well he has a book out now on the same subject.
I haven’t read it yet, but the table of contents includes good projects like:
Each section includes how-to, supplementary data, and a field notes page.
In most people’s minds, the recent financial crisis has been the last nail in the coffin of the Chicago School of Economics’ philosophies. That’s certainly what members of the saltwater school think, and I think that a lot of the public would agree with them. But what do they think in Chicago?
John Cassidy went to talk to some of them, and he wrote about what he found in the New Yorker (overview, Posner, Fama, Cochrane). The short version is that they don’t think that it proves anything of the sort. In fact, some of them even deny that there was a financial crisis at all.
Jonah Lehrer had a blog post yesterday about why this is what you would expect. You should also check out his recent Wired article about the neuroscience of learning from failure and Kevin Dunbar’s field studies of how scientists actually do their work.
Lehrer says that the reaction of the freshwater economists to recent events is a good example of how the discovery process really works. It also illustrates that you need to be interacting with people with diverse viewpoints to be able to benefit from failure. Because the members of the Chicago school only talked to each other they interpreted new information in terms of their theory, rather than modifying their theory to reflect new information.
Here’s Paul Krugman’s take from last September. As for Paul and the other saltwater economists, maybe they can help get the freshwater’s through the anger stage and into bargaining. But as Kübler-Ross could tell you, it’ll take some time.
With apologies to Elisabeth Kübler-Ross and Thomas Kuhn.
This is a short, computer generated film by Alex Roman.
Kahn's Exeter Short Film from Alex Roman on Vimeo.
It’s very nicely done. You might want to click through for the high-res version.
He did a longer piece called The Third & The Seventh.
I enjoyed this diagram from FlowingData.
Unfortunately the kids have been sliding down the left side just as I’ve been climbing up the right side.
This is hilarious. They are cupcakes themed after various games. See how many you can guess. Remember that the games have some video games, but not all of them are.
Click on the cupcake to take the quiz:
Simon Johnson has an article in the NYTimes today with another suggestion on reforming the finance industry. He compares the dangers we face from this industry to the dangers Eisenhower warned about with the defense industry in 1961. His conclusion is that we need to create something like a Lincoln Labs for finance. The idea appears to be based on this article about breakthrough ideas in the Harvard Business Review.
It’s kind of an interesting idea. One of the reasons we were given for deregulating the financial industry is that nobody in the government can really understand what the banks are doing. But the same is also true of the relationship between DOD and the companies who make the fighter planes and missiles, so perhaps there are some useful models there.
There are a bunch of other neat ideas in that HBR article, so you might want to read the whole thing.
On a related note, I just finished reading Elizabeth Warren’s book The Two-Income Trap. It also contains a number of interesting ideas about the recent debt crisis. She did a study which seems to show that the root cause of the American middle class’ increasing debt load wasn’t plasma TVs and SUVs. It was fixed costs resulting from a bidding war for access to quality education for our children. I recommend this book, but it’ll probably make you angry.
Timothy Egan had a cute article yesterday about our former governor. It’s about how Romney was bragging about how great “his” Massachusetts health insurance plan was when he was running for president last year, but is now taking pot shots at the nearly identical Senate plan.
It’s awfully easy to make fun of politicians for this sort of thing, but this one’s kind of funny.
On December 31st, Fabrice Bellard (founder of FFmpeg) announced that he’d calculated pi to 2,700 billion decimal places. Check out the details here. The coolest thing is that he did it using a single processor PC. Instead of throwing lots of expensive hardware at the problem, he worked on optimizing the software side. For very large variable precision numbers I/O becomes the bottleneck because you can’t keep them in memory. The run took a total of 116 days. He provides plenty of juicy details here and some samples of the digits here.
In related news, HP appears to have gotten a patent on one of the basic techniques used in this sort of computation, even though Knuth describes it pretty thoroughly in the second volume of The Art of Computer Programming.
We had our annual New Year’s day brunch today. Lots of people turned up. Lots of teenagers,
A few little kids,
and lots of old friends.
There was tons of good food as usual with our old classics (baked french toast, and cheddar sausage strata), as well as lots of yummy treats brought by the guests – too many to mention. We enjoyed getting a chance to catch up with everyone.
The kids played hours of four player Super Mario, but also found time to get out their instruments to play some music and chase each other around the yard with marshmallow guns and snowballs.