Looking Out To Sea

Check this one out:

Thanks for the CD, Liz! And now for some Bach:

I finished this at least a week ago, but I’ve been having a hard time putting in to words what I want to say. The book is The Man Who Mistook His Wife for a Hat by Oliver Sacks. It’s a collection of case notes about people he saw over the years in his work as a neurologist. Straightforward enough.

You must read this book.

That’s really all I need to say, though you probably feel a bit more explanation is required. Essentially, the book is a testament to the fragility of our minds. Most of the stuff we understand about the mind (hell, about biology) comes from what we can learn when it breaks. This book is about all the ways it can break.

• There are people who lose all sense of “ideal” entities. The man could hold a glove but only recognise that was soft cloth, that it was a container, that it had five little pouches.
• There are people who lose all sense of their body. The woman did not know what her limbs were doing when she wasn’t looking directly at them.
• There are people who lose all sense of inhibition. The woman would involuntarily mimic every person she saw as she walked down the street.
• There are people who lose huge chunks of their past, or who continually lose the immediate past.

If ever there was a book that showed the mundane, tangible nature of the brain, it is this. Everything that you can think of, every ability you have, is centred somewhere in that bundle of neurons in your skull. And if some of them should fail you could lose your ability to recognise faces, to recognise your own leg or to even understand the concept of “left”, the opposite of “right”. Can you imagine that?

Two stories popped up on my radar at about the same time, both about Facebook and social software. The first, by Clay Shirky, is a transcription of an excellent talk about [group dynamics on- and off-line, and why communities need to be protected from themselves] [enemy]. There’s a lot in there that seemed familiar and self-evident, but at the same time I had never thought about. That was also one of his points — that we have probably all seen communities die in the same ways over and over again, but no-one is learning from this.

The other story is — to put it politely — a slightly unhinged rant in The Guardian about [the dangers of Facebook and its ‘terrifying’ controllers] [friends]. It’s not often you find yourself in the company of a bona fide conspiracy theorist, and it’s difficult to know where to look. Reading Tom Hodgkinson’s article is a bit like that. You know if you looked him straight in the eye you’d probably just burst out laughing.

[enemy]: “A Group is its own Worst Enemy — Clay Shirky” [friends]: “With Friends like These… — Tom Hodgkinson, The Guardian”

Continue Reading »

First, solve the problem. Then, write the code.

— John Johnson

A few days [I looked at the game of Boggle] [previous] with a view to writing a program to solve the same problem that the user has: find as many (long) words as possible from a given grid of letters. I didn’t think much about how to implement it, just how to decompose the game into several discrete sub-problems.

The next stage is to think about how the problem might be solved conceptually — what approaches to take to structuring data, and how this might lend itself to an elegant solution.

[previous]: “Simple analysis of the rules and gameplay of Boggle”

Continue Reading »

There are good exhibitions and there are bad exhibitions.

On the first of January we visited the Basil Spence exhibition in the Dean Gallery. It was huge, comprehensive and fascinating. It was too big, in fact. We didn’t have the stamina or the time to see the last room (about Coventry Cathedral). There was lots to read, and it was all displayed with plenty of background information and interesting pieces. There were drawings from competitions he entered, drafts of school work, snippets from letters and television and radio.

I had never heard of him before but it was all incredibly interesting. It turns out he was the designer of George Square Library in Edinburgh — for which I don’t think we can forgive him. But he also designed some lovely fishing cottages in Dunbar.

Over the weekend we saw the Kylie exhibit at the Kelvingrove Museum, which was the very opposite in almost every way. A paragraph of introductory text at the door said “We hope you enjoy the story so far…” — but didn’t condescend to tell the story. You were expected to be intimately familiar with all her songs and albums (because it didn’t tell you anything about them), to know all about her as a person (what little text there was seemed always to start “And of course…” or “We are all familiar with…”). It was all contextless and entirely narrative-free.

There were amazing and intricately designed dresses pinned up everywhere — but no word on why they were designed, what the inspiration was, how they were used, whether Kylie Minogue enjoyed wearing them, etc etc. It was like looking through a clothes shop. The only thing I actually learned while I was there was that she had played the part of some kind of mechanic in Neighbours. I can learn more than that from Wikipedia, which I don’t need to queue for!

Finally, we also saw a small Quentin Blake exhibit at the Kelvingrove. Again, this was appallingly displayed. It was literally a rectangular block covered in frames, and a poster on the wall advertising that there was a Quentin Blake exhibit on. Not even a mention that you’re standing in front of it. It was the same poster as appeared at the front door downstairs. Really disappointing; thankfully the pictures were nice.

This is an official announcement of Helen’s new blog, The Sacred Art of Eating. She’ll be documenting a year’s worth of cooking challenge — to cook everything in Nigella Express before the 31 December.

Go and have a look!

I admit to being one of that set of users who prefer text-based email to HTML. This preference also shows in how I write my blog posts (in a text editor) and how I write letters for print (again, in a text editor).

The revolution will not be badly formatted

I can’t be sure what it is I prefer about the raw text. One facet of it is the sheer bloody-minded inadequacy of WYSIWYG tools. They are so prone to making deep, ugly nests of mutually exclusive formatting. With HTML editors this stuff is obvious:

<font size="2">A font size can be
  <font size="3">enlarged and then
    <font size="2">made smaller</font>
  </font>in really inadequate ways.
</font>

but that doesn’t mean it’s not obviously the cause of so many silly problems people have with word processors. “Why does it change font when I press return?!” or “Why is it writing in italics?” and so on.

I know that if I don’t ask for italics, or anything else, I won’t get them. This is quite satisfying. There will never be bits of invisible formatting floating across the page, waiting to ensnare my text like Spectres from a demon dimension. I can see everything.

Leave me alone with my art, dahlings

There’s also something untainted about writing without style or formatting. The literal placement of marks on a page, without thought to how they might look, is quite liberating. There’s a certain appeal to that, I can’t deny.

It also comes in really handy if I have to write code or similar. I would go to horrendous lengths to make sure smart quotes and ellipses and proportional-width fonts were all turned off. I read some programming mailing lists and the occasional HTML message really throws a spanner in the works.

Maybe one of these days we’ll have advanced enough user interfaces that can cope with all of these demands — slick styling, exact control and ease of use. But I’ve never met a good one yet.

As mentioned before, I have been thinking about solving Boggle. The best approach is always to start with what you know. We must first decompose the game into its discrete steps so that the problem doesn’t seem too overpowering.

These are my initial thoughts on the game, which may change as the implementation takes shape.

The Board

Boggle takes place on a 4x4 grid (I’m ignoring the 5x5 variant for now). Each place on the grid contains a letter, chosen randomly from six at that position. (In real life, these six are stamped on the faces of a die.) Most of the dice contain only single letters, but there’s also a “qu” variant to allow creation of legitimate English words with the letter Q.

The “qu” variant is a pain, because it means that there is no uniform representation of the grid contents. They’re not all single letters. But representing them all as strings seems wasteful because most of them are single-character strings.

The Game

The idea is to search for words that use adjacent letters on the randomised grid. You can’t reuse letters, but you can move horizontally, vertically and diagonally from any starting point. If you’re lucky you can spell a 16-letter word. In fact, if you get “qu” you can spell inconsequentially. That’s seventeen letters. ;-)

In my mind I can almost see the tightly-bound grid unravelling and hanging out into an n-ary tree. The n in this case is somewhere between 1 and 8, depending on how many neighbours a particular letter has. If only my animation skills were up to the task of showing it!

The words have to be at least three letters long, although this isn’t immediately helpful as far as implementation goes. It doesn’t suggest any useful shortcuts.

The End Game

Once you’ve got a word you have to check if it’s a real word. (This step is somewhat glossed over in real life because people shouldn’t just be writing down arbitrary collections of letters.) The idea is that you get points for long words, but if another player has the same word as you then neither of you get the point. Points are only awarded for unique words.

The person with the most points by the end of the match is the winner. The computer would not be an interesting player in this case. What would I do to simulate imperfect players?

• Put a limit on the character length. Long words are harder than short words for humans.
• Reduce the number of words in the dictionary, since not everybody has a huge vocabulary. Again, a bias for short words would be more realistic here.
• Implement a timer and search in a slow, random fashion so that not all the words are found on each run. Real people don’t work systematically through all the combinations (or maybe I’m just a poor player…).
• Some advantage to the player who can read the grid in the normal orientation, rather than from the side or upside-down.

If I randomly set some of these variables at the start of a round then it might be a more realistic game. Any other ways that the computer could be handicapped to be more like a human player?

Photo from 4rank’s Flickr page.

The fault is all mine. I have started learning more about networking, and so I want to start talking more about networking. But most people, they don’t know about networking. So there’s only one way to solve this problem. Order them to order large books from Amazon! Okay, that’s probably not very effective. The other way is to write about networking myself.

Consider this either a simplified introduction to networking, or a means of cementing my knowledge. You don’t properly know it until you’ve explained it to someone else. This does leave me with a small problem: I can’t actually tell if you’ve had it explained to you, or just told to you. So if there are any unclear bits please let me know.

Continue Reading »

Distractions don’t come more idle than looking up the TV Tropes wiki. It’s such fun.

The page on the stereotype of Hollywood atheist is quite interesting. It gets to the core of the “bitter atheist” — you know, the one that used to believe in God but doesn’t any more because his wife died in a car accident? It has this to say of Battlestar Galactica (the new series):

…features two prominent atheist characters, both of them wildly different: Admiral Adama, who views humanity as flawed but inherently good, and ultimately accountable to nobody but themselves for their mistakes in life, and Gaius Baltar, an egocentric technocrat who ultimately comes to consider himself a god.

Interesting summary. I haven’t seen enough BSG to be sure, but I’d always pegged Adama as being on the wishy-washy liberal theism fence. You know, Church of Scotland rather than Church of White Jesus From Texas. They forgot one character though — President’s aide Billy, who died in the very next episode after explicitly saying he was an atheist. But that’s what happens when you enter politics looking about 14 years old…

It is sad that the one atheist character who’s super-intelligent, a media personality, a hit with the ladies and a good-humoured guy also happens to be out of his tiny little mind. But you can’t have everything, right? ;-)

More interesting to ask, why are there so few positive role models of scientists in film and TV? The balanced scientists are as few and far between, and there is a lot of cross-over: the cold, calculated, “logical” scientist who can’t understand/engage with human emotion.