January 2006 Archives
The inevitable has happened. Actually, it happened a few months ago, but I felt compelled to write about it only today. I started getting spam in my Gmail account.
Actually, Gmail does a remarkable job filtering out the spam - MUCH better than Yahoo mail, and so far no false positives, which is much more than one can say for Yahoo.
That said, I've grown to rather - dare I say it? - enjoy perusing my spam folder on occasion, just because of the sender names. Today's highlights were Abolitionist R. Gravel and Glacier P. Stupendous. Do they think people will be fooled by these names? Or is their aim to try to compete for people's attention with humor? What other ridiculous senders have you seen in your spam inbox?
Ever since seeing reruns of the poker tournaments on TV in America - on the Travel Channel, of all places - I've been very excited to play. I downloaded a poker client a couple of weeks ago and I have been playing 7-card stud off and on every week. Last night, however, Heather and I tried, for the first time, a no-limit hold 'em game. Man, what a good game that is. So dramatic! We ended up winning big time, but then we lost it all in the end in a big bet. We weren't playing with real money, of course - but it is possible, with as low as four-cent (or is it ten-cent?) bet increments.
One of the really enjoyable aspects of the game is the trash-talking you can spew to your opponents. For each player, it shows his or her location, and a couple of the players were saying things like "go back to Tokyo - this is an American game!" which - I think - were in a spirit of lighthearted jest.
In any case, if anyone is ever up for some hold 'em, let me know. Heather's and my username on PokerStars is 0x0000029A.
The Museum of Bad Art. Need I say more?
Nowadays, advertisers do their best to target potential consumers by age bracket, race, gender, wealth, geolocation, and other data. Dish washing detergent and diaper TV commercials are placed in time slots coincident with television programs that appeal to women. Match-making services and Girls Gone Wild commercials are aired most often when lonely male geeks are watching TV, particularly during late night Star Trek or Battlestar Galactica marathons, for example. I've argued before that targeting offers not merely reflections, but also projections: it reinforces impressions and stereotypes within the viewers (you mean, Girls Gone Wild is not an accurate, positive reflection of women???).
Let's jump for a moment from this topic and talk about the increased sophistication of computers and home video gaming consoles. The XBox, with its XBox LIVE service, offers instant and significant connectivity with databases and other XBoxes all over the world. I'm sure that you can browse the web with your XBox - hell, that functionality is present in the PS2 and has been around since at least the Dreamcast.
A couple of years ago, Sony developed and released a technology for their PS2s called the EyeToy. It's a pretty fun deal; you can whack at virtual enemies floating in the air with your hands and the PlayStation will respond. You can do any number of really cool things with it, most of which has not yet even been discovered, likely.
The convergence of these kinds of technologies will result in some serious social paradigm shifts. The EyeToy has enough sophistication that it can determine what direction you're flicking your hand in; it is only a matter of time that it, or technologies like it, will be able to determine color, determine spacial arrangement, perform facial recognition, and more. It will be able to determine our race. It will be able to determine our gender. It will make judgments as to your age. It will read what we have sitting on the coffee table. It will be able to identify all the products within its range of sight and extrapolate your "socioeconomic status." Soon, a sophisticated microphone will be built in, and it will be listening to and extrapolating from our conversations. And our games and other forms of entertainment will naturally be the better for it.
Our consoles and computers will then connect with other machines and, in concert, determine what content will appeal to what individuals. Person A lives with his parents, eats a lot of Fritos, has a copy of Maxim lying on the shelf, and watches a lot of Sex and the City. Person B has similar circumstances; perhaps we can recommend that program to him? And while we're at it, let's push these similar products to him because there's a high chance that they will resonate with him.
But, of course, it's more than mere resonance. Our computers and consoles will become the primary agents of change for the way we speak, the things we buy, the people we meet and the things that we think. They already are agents of change; it's only going to get more intense, and, in the long run, it's inevitable.
Notice that I don't say worse, but more intense. I don't necessarily think that these are bad things - in fact I think that the human experience will be all the richer - but I do think that these are things about which we should be aware. At every turn, we should be vigilant about how we are steered by our technologies and information. How do we go about doing this? Well, for one thing we can start interacting with these marketeers so that, whenever we feel the urge, we can access something with a fresh perspective. A random article, if you will, but for the entire universe of information, entertainment, products, and services. This is just a first step, but an important one. Advertisers will work with us, if we want them to, to promote intellectual diversity in our society.
The scariest part about all this convergence of technology is misuse by the government. Fortunately, thanks in part to Mr. Gore, there is vociferous and prominent dissent to this latest privacy controversy. I don't know how effective the dissent will be in the short term, and I suspect that Washington will always spy on its citizenry, irrespective of the legality. Nevertheless, the convergence of technology and the ever-growing penetration of the Internet is incentive enough for us to be watchful of these trends.
Fascinating Radio Economics podcast. It touches on biology, evolution, psychology, and economics. Russ Roberts has an interesting comment about analyzing processes from an economics perspective. He characterizes this perspective as an exercise in analyzing incentives from market forces, not necessarily Wall Street market forces, but markets in a more abstract sense. Very insightful.
I found this book on the Open Book Project, and it looks to be a very nice intro to CS. It uses all C++, as well, which I really like
| Input image: |  |
| Generation 0: random image |  |
| Generation 1000 |  |
| Generation 10000 |  |
| Generation 200000 |  |
- Make the evaluation function lighter-weight by only sampling a random subset of pixels (this is closer to what happens in nature, so I think there will be no problem). This will make the program run faster.
- Use actual GP - e.g., instead of each element coding for a pixel, each element would code for a draw rectangle method. Fewer instructions will result, and possibly more interesting behavior.
- Extend the evaluation function by running it through a Sobel edge filter to detect edges.
- Extend the evaluation function by evaluating against not one, but several input images. Thus we have a multiple-exposure impressionism effect! (theoretically)
At the urging of Melanie, I performed a lexical analysis on a speech given by former president Clinton. I decided to perform the analysis on his 1999 State of the Union address, a transcript of which can be read on CNN's website. The results were fairly interesting, but nowhere quite as interesting as Mr. Bush's speeches. This is a qualitative and admittedly a subjective analysis on my part, but Mr. Bush's level of often-repeated rhetoric does not approach that of Mr. Clinton's.
Some highlights of the analysis follow. The phrase we will begin to meet our generation s historic responsibility was said three times in the speech. all americans can be proud that our was also said three times, as were i hope you will support it and all states and school districts must.
Mr Bush's longest, most often repeated phrase, setting an artificial deadline would send the wrong message to the, is repeated three times in his speech. This phrase is eleven words long. Mr Clinton's longest, discounting insignificant variation, we will begin to meet our generation s historic responsibility to, is ten words long, discounting artifacts like generation s, which should be generation's. However, Mr. Clinton's speech is about twice the length of Mr. Bush's. What conclusions can we draw from this? I can't be sure.
I'm really salivating over the prospect of analyzing the text of the press conferences held by Scott McClellan at the White House - updated daily. What morsels, I wonder, can we extract from those? Or do you think that this whole exercise is a waste of time and effort?
I have placed the complete output of the analysis in the extended entry, for any interested parties.
Does anybody know of a resource that explains, simply, the DDS file format, especially with regards to compression? I'm creating a tool that needs to read DDS files, with a minimum of DXT1 compression (at least I think that that would be sufficient), but I have no idea how to actually expand DXT1. Anyone know?
As I side note, I really want to avoid creating a D3D surface or using any OpenGL extensions. We're using an internal format, and it will be lightest and easiest for us to convert by hand directly from DDS to our format (unless we can convert first to BMP, then to our format, but that's a bit of a waste). Anyway, any help would be greatly appeciated!
I want to plug yet another new podcast that I just started listening to. Military History Podcast has quickly become one of my favorites. I have only listened to five episodes so far, but they've all been extremely compelling. In one cast the narrator talks about the whens, hows, and whys of using dogs as weapons in wars. Another deals with the role of valor in war in the ancient world. Yet another, 11 Unique Warriors, is a treatment about, for example, samurai, hoplites, berserkers, and the Praetorian Guard.
The most interesting thing I've learned from the podcast is the prominent role that the Greek Phalanxes had in promoting democratic ideals in the ancient Greek world. I'll leave it to you to listen to the hows and whys of it all. Really, really interesting podcast.
While I was home, I - as usual - watched a lot of movies. One of them was Open Water. Based on true events. Talk about the feel-good movie of the year!
Another film about which I had heard a lot, and finally got around to seeing during the holidays, was Crash. This film is interesting because it really underscores the racial tensions still present in modern America. Each of the many main characters undergoes a change in perception in how they interact with people of other races. Sometimes the characters become more tolerant in their attitudes. Other times, characters who arrogate about their progressiveness discover dark secrets about themselves brought on by racially-charged situations. Though the film is a wee bit simplistic in that all major characters transform dramatically, this certainly can be forgiven. It's Hollywood, after all! Great film, though.
I naturally watched a lot of TV while I was home, as well. After being away for eighteen months, some of the best, most interesting TV that I saw were the ads. Ads - in the U.S. and in Japan - are remarkably divulging of perceptions about cultural ideals. One ideal that struck me as remarkably forceful in its assertiveness had to do with race.
Commercial projections of the interaction of race and relationships struck me early. Every commercial I saw - and this has probably always been the case; I'm only now realizing it - sought to propagate the impression that people dating/marrying someone of their own race is the way things ought to be. A home-flooring commercial starring Bob Vila featured an all-black couple and then an all-white couple. A Capital One credit card commercial projected images of an all-white family. A computerized dating service was the most blatant; black with black, white with white, asian with asian. The message is clear: this is the way it always has been, and should be. Uniracial couples are the norm. Interracial couples are the exception.
Don't get me wrong. I can't really fault these companies for projecting these images. The fact is that the number of interracial couples in the U.S. is, on a percentage basis, fairly small - just under 3% according to a census report in 2002. In this light, one could correctly argue that these organizations - commercial enterprises - are projecting images that are accurate representations of reality. It is in the best interest of these commercial enterprises to project images of the status quo, because they want to maximize their appeal to potential customers. Economic forces - that is to say, companies' attempts to maximize value - thus shape the way agents in the commercial sector project their images onto the public. Sometimes, in the interest of profit (not a bad interest, I hasten to add), these images reinforce negative stereotypes. In some cases, the stereotypes might even seem benign on the surface, but they still seek to reinforce the idea that this is how you should be. This is what everyone else is doing. I'm thinking in particular of the Adidas (or was it Nike?) commercial in which a young black man runs through urban streets, encountering only other black men. He runs through and disrupts a dice game, and the participants, in apparent anger, up and chase him. Ahh, but he's got the shoes! He'll outrun them! He is prepared for the tribulations of urban life. You should be too.
Let's examine some stereotypes projected by this ad:
- The inner city is populated solely or mainly by black people.
- There are enough dice games played mainly by black people in the city that sometimes you have to run through one in order to get where you're going
- Black guys can run! Especially if they have the shoes.
- Black men are easily angered
Of course, these are my perceptions of the projected stereotypes, not necessarily the stereotypes that these companies intended to project. These are not, I hasten to add, reflections of my personal stereotypes, but rather my objective interpretations. So far as that's possible, which is probably not that far. I would be interested to hear what others think about it.
The other day I found a great blog about personal finance, entrepreneurship, and general effectiveness: iwillteachyoutoberich.com I've already added three books to my Amazon wishlist from the site. Very cool.
Wikipedia is still undergoing its fundraiser. I thought initially that this was in order to mount a good legal defense from the ridiculous Wikipedia Class Action Lawsuit, but this is mentioned nowhere in Jimmy Wales' personal appeal for donations.
This lawsuit against Wikipedia, I think, might be really damaging - perhaps fatally so - to the Wikimedia foundation. If you have some spare cash to donate, I recommend earmarking it for this valuable organization. Wikipedia has really been experiencing some growing pains lately; a year ago it was fast and easy to find articles; now, probably due to extremely heavy traffic, it takes forever.
In any case, if you've never heard of Wikipedia, I definitely recommend wasting a few hours of your time to peruse it.
My shoulder is killing me from doing too many drunken ninja rolls in the streets of Roppongi during Matt and Andrew's birthday party. Ouch!
The party was fun. It was kind of a roving party. I joined a little late due to my jetlag, but we ended up stopping at four (five by some reckonings) places. We also finished the night at Matsuya, a kind of Japanese-cuisine fast food joint, featuring mostly curries and donburimono. Think Yoshinoya ++.
I am a weirdness magnet, though. Earlier in the night, some guy came up to me and tried to catch my eye. He asked me something, but I can't remember what he said. It was hard to make out what he was saying, but it was definitely weird. Andrew said - and I don't know if it was in jest or not - something along the lines of "That guy thought you were a prostitute!" Later in the night, on our way to Matsuya, we came across a stability-challenged, very drunk white guy. I think he was English. He staggered up to me and said, "Sumimasen. Where's the chicken man?" Lots of shiny red blood was flowing out of his hand - it looked like it had been hit by a broken beer bottle or something. Fortunately, I knew all about the Chicken Man, though I think I was the only one, and the other members of our party, I think, thought he was crazy. I directed him to approximately where I thought The Chicken Man might be, but he went into a washoku place instead. Weird night!
When I told my Japanese teacher that I collected business cards, she soon started bringing me some of the best business cards that are in my collection. Truly, Japan is one of the best places in the world for high-quality, well-designed business and name cards. Take a look at this one:
This small image doesn't really do it justice, so I've included a larger version here. Apart from the expert use of negative space, of particular concern is the quality of the paper. The texture has been constructed in such a way that it scintillates as you reorient the card.
Here's another card from Kojima-sensei that makes great use of negative space:
Interestingly, both of these businesses are restaurants.
Last month, I had the pleasure of visiting Gonpachi, which is the inspiration for the restaurant featured in Tarantino's Kill Bill where Uma Thurman's character slaughters all the gangsters of the Crazy 88. The business card of this establishment is really something:
|
|
Shown above are the front and back. I didn't really get the contrast levels identical, so a little of the effect is lost. Nevertheless, what effect! The stylistic plant drawn in the background is subtle and beautiful. The logo's size and positioning, on the right side of the card, is again oriented in such a way that foreground negative space is put to good effect. The logo itself connotes a kind of humility and a sense of invitation. I only wish they had used a different typeface at the bottom.
The reverse of the card is on the right. The mini brand-type logos Kushi/Soba and Sushi really draw the eye and make a clear semantic separation between the two cuisines. It's absolutely brilliant. The map, at the bottom, is clear and simple. Best of all, the map is drawn all the way to the bottom and side edges of the card, lending the holder a sense of continuity - that this is not merely a place, but a special place amidst the large Tokyo map of establishments. Or so I feel. I'd be interested to hear what you think of this card.
I used to be extremely picky about the quality of cards that I acquired. That is to say, I only wanted cards that were in "mint condition." Well, recently I had an attitude shift, and I've started picking up old, trampled, discarded cards off of the street, and in train stations - to the undisguised disgust and contempt of some of the surrounding people. I picked this card up in Roppongi:
Either this is for a restaurant, of for a cooking school. The characters could be rendered, in english, as follows: Korea student town food/dining. It's in Okubo, near where I work. I'll have to check it out sometime. The written portion means from nine o'clock on.
Some cards I pick up are not as savory:
Well, that's really a totally unfair characterization. Salon de Trans is, well, let me try to do a translation of the text of the card:
A secret party for ??? and men! Cross-dressing, fetish, SM (sado-masochism), Bisexuality. Be excited by the fashionable. (or possibly, excite via the fashionable) And (possibly) a sweet night?
It seems that every business in Japan has a business card. In Kabukicho where I work, I have found some real choice cards:
Club Ego (I love that name - so appropriate) and Club Fantasia are very likely either so-called hostess bars
or brothels outright. Apparently the turnover is so high that the women (and men) who work in these places don't even get their name printed on the card; instead they write their names in (I mosaic-ed the names out to make them illegible). This is the case for all of the club cards that I own. Another reason the cardholders write their names might be that they are trying to endow the cards with a sense of intimacy. There is support for this argument by the fact that the phone numbers and emails of the girls are hand-written on the back of the cards. In the case of Club Fantasia, a more personalized note is written in the upper-left hand corner. It congratulates someone by name and then urges them to がんばってね, which might be translated as hang in there or do your best. Just an extra bit of personalization, I suppose.
Of additional interest on that card is the image in the bottom right-hand corner of the card. Many Japanese cards and signs have images like this. It's purpose is to provide a fast web hookup from cellular phones. All cellular phones are equipped with cameras in Japan, and if you take a picture of the image, it will recognize the image and send your phone browser to the location associated with the image. In this case, probably the URL of the club. These images are also present on advertisement posters on the train. Wild, eh?
If anyone is interested in trading business or name cards, I would be thrilled to hear from you. I have duplicates of just about every card I own for this purpose.
I spent about about twelve days at home for the holidays, and it was great. While I was there, I helped set up blogs for my mom and dad. Check them out:
The WoodOhio Blog
This 'n That
My mom already has two entries up. Feel free to leave them comments!
As a software professional, I certainly have had some exposure to the concept of refactoring. But up until I read a volume on the topic by Martin Fowler et. al., the full importance of the art of refactoring really hadn't hit home.
An excerpt from the preface of the book:
What Is Refactoring?
Refactoring is the process of changing a software system in such a way that it does not alter the external behavior of the code yet improves its internal structure. It is a disciplined way to clean up code that minimizes the chances of introducing bugs. In essence when you refactor you are improving the design of the code after it has been written.
One might think, "what is the point of changing the internal structure - the source code - in such a way that no changes are made manifest in the functionality of the program?" It's like taking a piece of DNA and changing it in such a way that the phenotypes are the same before and after the change.
There are some good reasons that one would want to make such changes, however. One is to promote clarity. If a programmer cannot understand a system, he or she cannot add functionality to the system in such a way that unintended side effects don't rear their ugly heads - or at the very least, functionality extention is very difficult in large unrefactored systems.
The terminology refactoring is extremely appropriate, I think. It recalls algebraic refactoring. To wit, an expression such as (x² - 6x + 9) can be refactored to (x-3)(x-3), or (x-3)². The evaluation of the two expressions are identical, and yet the second expressions - especially when expressed as (x-3)² - are much easier for us to get our head around.
Software development is very similar to this process. Effective software development is all about developing and identifying existing safe, effective software primitives, and combinging these primitives in such a way that we develop something useful, and also in such a way that we don't reinvent the wheel. In terms of clarity, let's return again to algebra. (x² - 6x + 9)² is a combination that is semantically equivalent to (x-3)^4, but if we already know that (x-3)² is a factor of (x² - 6x + 9), then there is no reason, in terms of comprehensibility, that we should express it the lengthy way. (Now in terms of execution speed, there are some legitimate concerns about leaving a piece of code factored in decidedly non-primitive (non-prime?) terms, but most of these situations can be ignored until profile time, and the book makes convincing cases about when, where and how you should take these cases into consideration - namely, after refactoring.) Software development is a lot like algebraic expression construction in this regard. Instead of matematical functions and operators, programs are composited by procedural functions and objects.
I recently had a conversation with a former colleague about refactoring and software testing. I regard this colleague very highly, but still he is not yet convinced about these software refinement processes. As the head of a software firm, he is legitimately concerned about the allocation of software developers' time to testing and restructuring - in other words, to activities other than functionality extention. To him, I should have pointed out the fact that the vast majority of time in software development is devoted to determining sources of software defects, according to some studies (unfortunately, I would like to have provided a reference for this statistic here, but I don't know where I read it. My gut tells me that it was in Code Complete by McConnell. ) Finding software defects in a well-tested (I mean automated tests, of course), well-factored system reduces debugging from sometimes days-long events to minutes, in many cases. In essence, there will be more time left over to devote to the extention of functionality. That's an outcome about which I think any software developer would be excited.
The fantasy of software developers is that, when they begin a project, they will start fresh without having to deal with legacy code, and be able to implement the features needed exactly the way they want. The reality is that software engineers often inherit a complicated, sometimes unworking codebase. The techniques in this book offer a transactional roadmap - and by transactional, I mean that it is trivial to roll back changes - toward making your codebase more comprehensible. Among the book features:
- Explanation of refactoring
- Many good, clear examples
- A large technique catalog - over seventy, in fact. Think in terms of the Design Patterns examples by Gamma, et. al.
- Advocation and demonstration of automated software techniques
- All examples are in Java
Although in many regards a strength, one of the biggest problems I have with the book is that all the examples are in Java. Java is a great language, don't get me wrong, but it seems that some (but not most) of the techniques are geared towards Java in particular. In this respect, the book reeks of language evangelicism (talk about code smells!). If the authors wanted to support language-specific idiosyncracies, I wish they had examined several languages instead of just one (although Smalltalk is mentioned, it, as usual, seems to be relegated to the theoretical). William Opdyke does provide a useful, four-page excerpt about refactoring in C++ near the end of the book, but it's not enough.
Despite this very minor flaw - a flaw that certainly makes itself far more manifest in other software texts - Refacoring is a great resource - a must have - for serious software developers.
At long last, I have completed The Roads to Sata: A 2000-Mile Walk Through Japan - and when I mean at long last, I mean six years. During my first stint in Japan, my friend Brian showed the book to me. The first passages impressed me deeply enough that I decided to buy the book immediately. But it's not the kind of book that I could read cover-to-cover; it simply doesn't read like that. Quickly I put the book down and forgot about it until very recently.
Mr. Booth, an Englishman who had lived in Japan for seven years at the time of his writing the book, took a four-month walk from the northernmost point of the main four islands of Japan, Cape Soya, to the southernmost point, Cape Sata. The Roads to Sata is a record of the things he saw and the people he met. It's a fascinating set of insights into traditional Japanese attitudes, particularly in regards to foreigners.
As it turns out, I have had the opportunity to live in two of the towns that Mr. Booth visited: Naoetsu, which is now part of Joetsu, in Niigata prefecture; and also Komatsu in Ishikawa - although Mr. Booth doesn't mention Komatsu by name. Instead, he describes Ataka, a community that was incorporated into Komatsu perhaps a decade ago. Incidentally, this phenomenon of merging smaller communities into larger cities is extremely common in Japan, and is motivated by tax incentives from the national government.
Booth's description of Naoetsu is far from flattering...
Naoetsu is described in the official guidebook as "one of the flourishing industrial centers on the Japan Sea coast." It is so flourishing that from a distance of four kilometers you can't see it. As you get closer the mechanics of its disguise become apparent. The chimneys of the Nippon Stainless factory and the Mitsubishi petrochemical complexes pump a solid stream of choking brown smoke into the Sunday afternoon sky. The dock is full of cranes and filthy little tramp steamers, and a continuous trickle of dust filters down from the snow roofs that ward off nature from the pavements. Naoetsu has the distinction of being the only city in Japan whose beer shops I raced by without a second blink. I had a vision of petrochemical yeast dissolving most of my vital organs which were then replaced by a stainless steel liver and an injection-molded Mitsubishi stomach. On the billboards of petrol stations gigantic bikini-clad women profered cans of motor oil at thirsty motorists. I fled Naoetsu in top gear and didn't look back at it till forty minutes later by which time it disappeared.
Ouch! Naoetsu was no paradise, but even cynical I wouldn't frame it this scathingly. It definitely has a kind of charm. Within Niigata, there are also very interesting passages about Kashiwazaki, Kakizaki, and Itoigawa. There's a particularly poignant passage in Kashiwazaki.
As it turns out, the Ataka barrier gate near Komatsu is the scene of a very famous legend about Benkei and Yoshitsune. This legend, and legends like it, have been very important in helping to define traditional Japanese national identity - vestiges of which indubitably remain today (if there's any interest, I'll expound upon this legend). A conversation that Booth has with an old man at Ataka that impressed me deeply:
The old man gave me the directions I wanted to the site of the Ataka barrier gate. He also gave me a withering scowl.
"What on earth d'you want to go and see that for?"
"I'm interested in history."
He squinted down at the little cassette recorder that I carried in a pouch on my belt for making notes.
"You a writer, then?" he asked, perceptively (someone else had thought it was a digital brandy flask). Yes, I told him, I had written a few things.
"Well, just you remember this," he said, getting up off his stool for the conclusion of his lecture, so that he stood a good deal taller than I did: "A country is like a sheet of paper; it's got two sides. On one side there's a lot of fancy lettering - that's the side that gets flaunted about in public. But there's always a reverse side to a piece of paper - a side that might have ugly doodlings on it, or bits of graffiti, or goodness knows what. If you're going to write about a country, make good and sure you write about both sides."
The Roads to Sata really has little to do with character development or explaining the "inscrutable Japanese," whatever that means. Instead, it's all about Booth's impressions, largely (but not entirely) without judgments. It's a good read in bite-size portions, and it's extremely pithy, if you know what to look for. Booth also wrote a similar work called Looking for the Lost: Journeys Through a Vanishing Japan - a book that I'm really looking forward to reading in the future. Anyone read this yet?
