Dani's Blog

I've re-homed my blog and registered a nice domain for it:

www.danielfortunov.com

The content is managed and hosted by WebJam -- check them out and make your own web jam today, for free!

Let's face it, I've been lazily putting up with Multiply for too long. Clearly their focus is more on social networking than straight-up blogging. They have some nice features, but I can do better.

I'm currently evaluating alternative blogging solutions. My selection criteria, in priority order, are:

1. Effortless : I want a rich WYSIWYG editing experience. Having to hack HTML, even occasionally, is a major hindrance.
   
2. Free : Publicly available; no private hosting required.
   
3. Photo Gallery Support : So I can publish photos when I want to.
   
4. Anonymous commenting : People shouldn't have to sign up for an account just to post a comment.
   
5. Backups : I want to hold on to my content, in case the provider goes away, or I want to move elsewhere.

Suggestions welcome...

In the year that I was born, 1983, Stephen Hawking was interviewed by the BBC for Horizon, speaking through an interpreter who deciphers his severely slurred speech.



Twenty-five years later, the marvels of modern technology bring snippets of this footage to your desktop:




I've always admired Stephen, ever since I knew of him, for his positive outlook, his incredible intellect, his sense of humour, and his implicit demonstration of the power of persistence.

Where do the things we consume come from, and go to when we're finished with them?



Let Annie tell you, with her short instructional video: The Story Of Stuff.

Unfortunately, due to hard-drive failure, I am unable to bring you the super-geeky St. Valentine's Day joke that has been waiting patiently to be re-printed for the past two years. I guess it wasn't meant to be. Perhaps next year it will work out...

In the mean time, you can keep yourself amused by reading the works of Philip M. Parker. He has apparently "written" somewhere between 85,000 and 200,000 books in the past five years, since devising an automated mechanism for getting a computer to write entire books on-demand, in less than ten minutes.

Philip is clearly targeting The Long Tail, for his books have an exquisite niche appeal with titles such as The 2007-2012 World Outlook for Manufacturing Capacitors, Resistors, Microprocessors, Bare and Loaded Printed Circuit Boards, Electron Tubes, Electronic Connectors, Computer Modems, Semiconductors, and Other Electronic Components. This one-of-a-kind tome can be the highlight of your mantle piece for a mere 408.68 British pounds.

The Annals of Improbable Research have scooped the story in their two part article, appearing in The Guardian. (Recommended reading.)

Here is a short video that talks about Philip's book, game, and video automation system:

I'm just waiting for Chris Morris to jump out at any minute...

If you want a good introduction to the world of cryptography, check out this four-part Google Tech Talks series. The first three lectures are delivered by Steve Weis; the last is a guest lecture from Ben Adida, who covers the details of electronic voting systems.

These sessions were hosted weekly, by Google University, between 28 November and 19 December, 2007.

1. Introduction to Modern Cryptography (54 min.)
   Steve Weis, 28 November 2007
   
   Steve starts with the definitions and historical origins of cryptography, presents a number of cryptographic systems, and talks about modern symmetric algorithms such as DES and AES.
   
   He also covers the problems with symmetric encryption and then goes into the concepts relating to asymmetric Public Key Cryptography (which I have thought to be wonderfully elegant, ever since I read the extensive documentation that came with one of the early versions of Phil Zimmermann's PGP).
   
   
   
2. Cryptography in Practice (59 min.)
   Steve Weis, 5 December 2007
   
   In this lecture Steve talks about the difficulties of creating strong cryptographic algorithms, the different types of algorithms and their uses, and the practicalities of  correctly assembling these algorithms into secure system. His content reminds me of Bruce Schneier's book Practical Cryptography, which I highly recommend for anyone interested in the subject.
   
   It's incredible what some people come up with, such as an attack on RFID encryption that analyses the minor variations in power consumption of the device to deduce what it is doing internally!
   
   
   
3. Security Proofs and Definitions (49 min.)
   Steve Weis, 12 December 2007
   
   Steve delves into the detail of proving the security of a cryptographic algorithm. Topics covered include semantic security, the chosen-ciphertext ("lunchtime") attack (CCA), it's older brother, the adaptive chosen-ciphertext attack (CCA2), and zero-knowledge proofs.
   
   This is the sort of stuff that people should be forcibly educated about every time they even think about implementing their own proprietary algorithms.
   
   
   
4. Verifying Elections with Cryptography (92 min.)
   Ben Adida, 19 December 2007
   
   Electronic voting is a very interesting topic, mainly due to the scale of national voting and the conflicting requirements of letting an individual verify their vote has been counted correctly, but without allowing them to prove this to anyone else (e.g. someone who is bribing or forcing them to vote for a particular candidate).
   
   Ben gives in introduction to the problems involved in electronic voting and then presents his "Scratch & Vote" system for a cryptographically strong process that is completely transparent and publicly verifiable at every step of the way.
   
   Electronic voting is also one of Bruce Schneier's favourite topics, and he has previously drawn attention to another voting system which had a similar 'feel' to "Scratch & Vote".
   

Once you've watched all those videos you should be suitably prepared to find this comic amusing:


If you can't be bothered with any of that, then at least promise me that you'll read this article from last September's MSDN Magazine. I guarantee that you will be amazed with the amount of thought needs to go into something as fundamental as transforming the range of a single random integer... even in the cushy surroundings of the .NET Framework:

PS: There are also some good articles in the security-themed November 2007 issue of MSDN Magazine.

Nearly a year ago, this device was Exhibited at the "IDE Performance Show" (whatever that is).

OTTO (Greek for ‘ear’); a device that makes hidden sounds audible. This is achieved via a thin polymer piezoelectric contact that senses weak vibrations and plays them as a sound through an integrated speaker. OTTO can be positioned on almost any surface through a combination of suction and magnets. By placing several units on different objects, one can select and create a new sonic experience and a form of ambient music appreciation, thereby utilising our space as a multidirectional audio platform.

It's an interesting idea — I like the thought, and the form factor.

There's more information and lots of photos and design sketches on their blog. (Although it's a little out of date and the video links are no longer functional.)

Donald Norman demonstrates the significance of cultural constraints with a thought-experiment:

Next time you are in an elevator, stand facing the rear. Look at the strangers in the elevator and smile. Or scowl. Or say hello. Or say, "Are you feeling well? You don't look well." Walk up to random passerby and give them some money. Say something like, "You make me feel good, so here is some money." In a bus or streetcar, give your seat to the next athletic-looking teenager you see. The act is especially effective if you are elderly or pregnant.
            — Donald Norman, The Design of Everyday Things (pp. 86)

It is interesting to note that such action would likely have you labelled as crazy by the general public, when in fact all you are doing is violating their in-built cultural constraints.

(Image Source: TorrentFreak Anti-DRM T-Shirt Design Contest)In the financial world (which is roughly equivalent to gambling) the acceptance of risk necessarily equates to financial returns. The bigger the risk, the bigger the returns.

Think about roulette odds: Put your chips on Black or Red and you have a 53% 'risk' of losing, and a 1:1 pay-out for winning; put your chips on 00 and you have a 97% 'risk' of losing, but a 35:1 pay-out if you get lucky. The returns you get are commensurate with the amount of risk you take on.


What does this have to do with your shiny new iPod, I hear you ask? Well, the link is that whenever you buy DRM-protected music (e.g. from the iTunes Music Store), you are taking on certain risks without realising it, and without adequate compensation.

An article published last month on kuro5hin presents two of these risks as yet another argument against DRM:

Credit Risk

"you're counting on Microsoft or Apple not going out of business any time in your life and making all of your media instantly unreadable"

Liquidity Risk

"Unlike with books, CDs or DVDs, which can be given to friends for free or even sold on the newly viable secondary markets facilitated by eBay and others, DRM media is generally non-transferrable."

An Example

Suppose that you would like to buy the latest album from your favourite band. You are trying to decide between instant gratification from the iTunes Music Store or a slightly more arduous trip to the high-street. A quick check shows that prices between the two are comparable: £7.99.

A store-bought CD is a physical asset. It is something you can hold, and carry with you, or lend to friends. It had to be manufactured, and transported to the high-street. It has a nice printed cover and, if you're lucky, it may even have a fold-out booklet with song lyrics, artwork, etc. It can be ripped and encoded to any format you like, and played on a portable music player, or in-car player, or entertainment centre of your choice. It can even be sold at the flea-market if you're strapped for cash. A CD is not at the mercy of your iPod's declining battery-life.

Why, then, should you pay £7.99 for an inflexible, low-fidelity, risk-exposed, DRM-secured MP3 download, when you could have the CD album instead?

These considerations also extend beyond music to other forms of protected media. An example from the article is "Sony's PlayStation 3, which has the capability to tie the video games played on it to a unique machine" meaning that you can't lend or sell your game to anyone else.

(Image Source: EnergyPortal.eu)

In his blog, Chris Anderson presents a step-by-step breakdown of the carbon effects of printed magazines, then compares this to reading an electronic version of the material online.

From media hype on the matter of global warming you'd be forgiven for assuming that it is surely worse to cut down trees, turn them into paper, print magazines, and deliver them to readers with a fossil-fuel guzzling vehicle, only to have them be thrown away some time later.


In his post, Chris presents the somewhat counterintuitive conclusion that the "dead-tree" (printed) magazines actually have "a smaller net carbon footprint than web media".
        Read his full post: Are dead-tree magazines good or bad for the climate?

His analysis is merely qualitative, but there's also a more quantitative analysis in a paper (pdf) from the Royal Institute of Technology, Stockholm.

---

(Chris started this blog, entitled The Long Tail, to document the progress of a book by the same title. His book is now out, and on my reading list. Stand-by for a book review, just as soon as I get around to reading it.)

The folks at Carnegie Mellon University have come up with a novel way to engage students in Java programming courses by essentially turning Java programming into a game. They call it Alice.


I was skeptical at first but it seems pretty decent: essentially they set a more interesting context for the code that students write, by having it control characters in a 3D environment. This lets the students have a somewhat more engaging learning experience than they would starting off their course in the more traditional way of staring at a screen of code.


Check out the second video on this page for a nice demo of what it's all about.
(The first video is just marketing fud, consisting solely of people talking it up but without actually explaining what it is; stay away from the first video and go straight to the second one. Sorry, I couldn't work out a way to embed the video here.)

They seem to have had success with it on their courses. Oh, and it's free.

The Design of Everyday Things, by Donald A. Norman, is all about the little details in our everyday lives. Basic things like doors, or telephones, or stoves. The design details of these objects can either be done well, or poorly. but without conscious effort we don't notice them either way.

Ideally, the design details are done well. We don't notice them (and we don't need to). If done poorly, however, we still don't notice them; instead we fault ourselves: "I'm so bad with mechanical things." "I don't get on with computers." We naturally tend to fall into the pit of learned helplessness (or taught helplessness).

It's easy to fail as a result of poor design. Even something as trivial as a door has significant level of complexity when you break it down: Do you push or pull the door to open it? If so, on which side? Or maybe it slides? If so, in which direction? There are four ways to manually open a swinging door; how do we know which mode (or modes) of operation apply? The answer lies in the subtle design queues that answer these questions for us.

How many times have you pushed a door that should be pulled open, or pulled a door that should be pushed open? This picture shows a door giving mixed messages: The sign explicitly tells you to push the door open, but the handle implicitly tells you to pull the door open; because, after all, handles are for pulling on!

A classic example I can remember is one of the entrances to the Mathematics building at Reading University: There are two sets of double-doors with identical features on both sides. A horizontal plate positioned at hip-height and spanning the width of the door, suitable for pushing but also raised enough from the surface of the door so that you can wrap your fingers around the back of it and pull the door. At least it was obvious that the double-doors were meant to open in the middle, but the direction in which each set opened was a mystery to me. I failed at those doors many a time, and many others did too. (Don't even get me started on the light switches.)



Only yesterday I saw a man at a service station try to open a door similar to the one pictured by pulling on the handle. The sign clearly instructed to "Push"! Was he stupid? Should he have read the sign? What if he was not literate in English? What if he was distracted? Was it really his fault that he failed to open the door on the first attempt, or was it the fault of poor design?

"Somehow, when a device as simple as a door has to come with an instruction manual -- even a one-word manual -- then it is a failure, poorly designed."

"If a design depends upon labels, it may be faulty. Labels are important and often necessary, but the appropriate use of natural mappings can minimise the need for them. Wherever labels seems necessary, consider another design."

Another scenario Norman talks about in his book is the arrangement of knobs on a stove-top: Burners are in a rectangular arrangement, yet the control knobs for them are, more often than not, in a linear arrangement. Left and right pairs of knobs control left and right pairs of burners respectively, that much is intuitive, but which dial in the pair controls the front burner and which controls the rear is a mystery.

Labels and diagrams are sometimes helpful, but often mask poor design.

• Consider the benefit of a rectangular arrangement of the hob controls which mimics the rectangular arrangement of the burners. Would there still be need for diagrams that indicate the mapping of controls to burners?
• Consider the benefit of having a rectangular push plate on one side of a door, and a pull handle on the other: Would there still be a need for signs indicating "Push" and "Pull"?

Good design is intuitive and easy to use right out of the box -- without having to read the manual. After all, few people actually read the manual anyway. If your instructions aren't already blindingly obvious then, even if the manual is read, its content will soon be forgotten by all but the most dedicated users. Perhaps the most amusing demonstration that "nobody reads the manual" is the book itself: First published in 1988, this book is "the manual" for designers, yet it seems a lot of its lessons have gone unlearnt over the past two decades.

"Good design takes care, planning, thought. It takes conscious attention to the needs of the user."

Computer Software

The cunning thing is that all of these design principles also apply in the world of computer software:

The Pit of Success: in stark contrast to a summit, a peak, or a journey across a desert to find victory though many trials and surprises, we want our customers to simply fall into winning practices by using our platform and frameworks. To the extent that we make it easy to get into trouble, we fail. -- Rico Mariani

The Pit Of Success for doors: A door without a handle can only be pushed.

Another good example of how well this content relates to computer software is the principle: Design For Errors.

There are many different reasons why users may make errors. Norman names seven distinct classes of error, and prescribes that designers should:

1. Understand the causes of error and design in order and minimize those causes.
2. Make it possible to reverse actions – to "undo" them – or make it harder to do what cannot be reversed.
3. Make it easier to discover the errors that do occur, and make them easier to correct.
   
4. Change the attitude toward errors. Think of an object’s user as attempting to do a task, getting there by imperfect approximations. Don’t think of the user as making errors; think of the actions as approximations of what is desired.

The Windows Recycle Bin was a way to allow recovery from the previously permanent action of deleting a file. Even the "Are you sure?" prompt was not enough, for that tends to confirm the action (deletion) when what you really want to confirm is the file that is being deleted. Users tend to hit "Yes" with a sigh: "Of course I'm sure; I wouldn't have pressed delete otherwise!" only to realise a second later that they accidentally selected the wrong file.

Good designs will let the user take steps towards their goal and give them suitable feedback to allow them to reach it, but without imposing serious penalties or restrictions on the path they take to get there.

"The principle of good design can make complexity manageable."


(The content from this book is nicely summarised here.)

Here's a nice photo I came across on flickr, of the swimming pool at Coogee Beach, Sydney:

I can't take any credit for this shot, but the water does remind me of this photo I took last year:
November 2006: Trafalgar Square by Moonlight

I just ordered an upgrade for my current phone. I'm sticking with the Orange SmartPhone product line (which is merely a re-branding of HTC hardware) but taking an upgrade to the E650. It has a larger 2.4" screen, 2 mega-pixel digital camera, Windows Mobile 6, and a slide-out Qwerty keyboard. (Alas, they don't have a Dvorak option.)

Getting the upgrade was not as easy as it should be, partly because I'm on an O2 equivalent plan (through the Orange Value Promise offering) in an attempt to reduce the monthly contract charge for my phone, which I use fairly lightly.

Customer Service: Upgrades
I called Orange and the upgrades department told me my desired phone would cost £59. I told them that I'd been a customer for six years, and was now running two years out of contract, having taken no upgrades. I think after three years of "paying off my handset" with the monthly contract rates I'm entitled to an upgrade. The representative said it was based on my past six months of usage only. (So having only two handsets over my six contract years with Orange didn't make any difference!)

"You're telling me that after six years with Orange and two years being out of contract you want me to pay £70 for an upgrade? Ok, in that case I'd like to speak to someone who can help me switch to one of your competitors."

He warned that there was a 20 minute wait and offered to call me back (what, too many people trying to disconnect at the same time?! hah). Half-expecting not to ever get a call back I was pleasantly surprised when I did, only 10 or 15 minutes later.

Customer Service: Disconnections
The woman barely let me finish my spiel about how loyal a customer I am, and how one of her colleagues had just informed me that there was absolutely no way for me to get this phone for less than £69.99.

"Not necessarily. Let me see... Since you're a low usage customer we can offer you..."

She proceeded to detail a contract that featured monthly minutes and text allocations that were an order of magnitude greater than my current contract, yet for a cheaper monthly cost; and I'd get the E650 for free.

"What is this plan called? Is it listed on your website or anything?"

"It's called the Generic £15 plan; It's not listed anywhere; It's specially tailored, just for you."

(Yeah right, it's tailored just for me and any other customer who's about to leave. But hey, I'm not complaining!)

Now we're talking
I was hoping just for a free phone... and I didn't think it would be that easy either! The monthly cost reduction was a nice bonus, and the extra minutes/texts will ensure I never exceed my monthly allocation.

I get the phone I want, and reduce my monthly outgoings; they get to keep me as a customer. It's a win-win situation both ways.

Moral of the story
Drive a hard bargain with your mobile carrier, especially when it comes to renewals. Don't bother with the upgrades department; go straight to disconnections (internally named "customer retention") — they are the ones with the power to offer you un-advertised upgrade options and contract plans.

With the increasing maturity of the mobile phone market, providers will only get more competitive in their strategies to snatch customers from the competition, and to then retain them.

Following on from my previous pictures of bullied bikes comes this modern-art marvel. Spotted in a mini-park near Canary Wharf, resting upright with the forks jammed into the ground.

That rear wheel looks a little worse for wear, and the chain could do with a little lube.

Image from stock.xchngNot my idea:

You are placed into an arena devoid of weapons and structures. Once every two minutes, a raving mad starved five-year-old is released into the arena, bent on killing you. You must incapacitate the child, or given enough time it will kill you. At twenty minute intervals, the amount of five-year-olds sent into the arena is doubled. Assume the children are unarmed, possess no special training, and are in a state of blood-lust. You cannot take a weapon in. You must rely on your superior physique, endurance, and skill.

How long could you survive?

Presumably there is only a finite time before the arena is too full to accept further children, but I guess that only matters if you survive that long. Of course, in less than twelve hours the (theoretical) number of people inside the arena would exceed the world population, so there's also hope for the supply of children being exhausted, if your arena is really big.

As every man goes through life he fills in a number of forms for the record, each containing a number of questions . .. There are thus hundreds of little threads radiating from every man, millions of threads in all. If these threads were suddenly to become visible, the whole sky would look like a spider's web, and if they materialized as rubber bands, buses; trams and even people would all lose the ability to move, and the wind would be unable to carry torn-up newspapers or autumn leaves along the streets of the city. They are not visible, they are not material, but every man is constantly aware of their existence.... Each man, permanently aware of his own invisible threads, naturally develops a respect for the people who manipulate the threads.

     — Alexander Solzhenitsyn, Cancer Ward, 1968.

Source: Schneier.

I've mentioned this book with a couple of quotations in my previous posts. I've finished reading it now and I have to say it was a pretty good read. It covers a lot of the major things in his life, like going for walks with his father as a child, and making the atomic bomb; telling the reporters to call back later when they were informing him that he'd won the Nobel Prize, and all that other run-of-the-mill stuff. He also talks about some more wacky things — like how we should in theory be able to write all of the books in the world down on a grain of rice — and enters into philosophical discussions of science, religion, society, etc.

Somewhat confusing is that this book was first published more than ten years after the death of Richard Feynman. The reason is that it is actually a collection of his previous works (including transcripts from talks and BBC documentaries), edited together by Jeffrey Robbins, like a final parting shot from the grave.

And for my own parting shot, I leave you with this amusing quotation from his school days:

I couldn't answer them — which made me realize I couldn't talk while I was counting to myself — and kept right on running up and down the stairs, looking like an idiot.

The guys at the graduate college were used to me looking like an idiot. On another occasion, for example, a guy came into my room — I had forgotten to lock the door during the "experiment" — and found me in a chair wearing my heavy sheepskin coat, leaning out of the wide-open window in the dead of winter, holding a pot in one hand and stirring with the other. "Don't bother me! Don't bother me!" I said. I was stirring the Jell-O and watching it closely: I had gotten curious as to whether Jell-O would coagulate in the cold if you kept it moving all the time.

From It's as Simple as One, Two, Three pp. 217-223 in The Pleasure of Finding Things Out: The Best Short Works of Richard Feynman.

I was born on 31 January 1979 - a Wednesday. I know it was a Wednesday, because the date is blue in my mind and Wednesdays are always blue, like the number nine or the sound of loud voices arguing.

So begins Born On A Blue Day, by Daniel Tammet — An extremely well-written and engaging autobiographical account of growing up with Asperger syndrome.

Daniel is a savant in that he can do things like memorise the mathematical constant π (pi) to 22,514 digits and then spend 5 hours 9 minutes and 24 seconds reciting it to a panel of professors at Oxford University, or learn Icelandic from scratch in a week before giving a fifteen minute television interview in front of a live studio audience... in Icelandic.

The reason he can carry out these feats is explained in the opening paragraph of his book (reproduced above) — because to him numbers are much more than merely a set of digits. His brain associates each number up to 10,000 with a particular feeling, shape, colour, and texture. When doing something like memorising π (pi) these feelings and textures string together to form scenes, and landscapes, and an entire world for him to be immersed in. He even captured the landscape that he associates with the first twenty digits as a painting, which he recently auctioned off on eBay (for charity).

Having read some books by the likes of Oliver Sacks and Vilayanur Ramachandran¹, it makes perfect sense that his excellence in mathematics and languages are so well correlated. It has long been thought that the parts of the brain involved in natural language processing and those involved in mathematics are significantly, if not entirely, overlapped. The reason that most people don't take as kindly to mathematics as they do to language is that mathematics is more optional than speaking; they don't persist long enough to discover the beauty of it.

As Darren Aronofsky said, "Mathematics is the language of nature."

In addition to his crash-course in Icelandic, he is also fluent in French, Finnish, German, Spanish, Lithuanian, Romanian, Estonian, Welsh, Esperanto, and English. And in addition to memorising π (pi) to 22,514 digits, he can also tell you that 377 multiplied by 795 is 299,715, instantly. "When I multiply numbers together, I see two shapes. The image starts to change and evolve, and a third shape emerges. That's the answer. It's mental imagery. It's like maths without having to think."

Perhaps the most impressive thing about Daniel Tammet is the progress he has made since his childhood in integrating himself into the world around him, working to understand social interactions, people's emotions, and other things that most people 'get' automatically and take for granted. For Daniel, the ability to tell when someone is sad, through non-verbal cues alone, is as strange as the ability to reel off the cube root of 185793 is to us. Unsurprisingly, it was quite difficult for him to fit in amongst his peers through school, but it sounds like he's come a long way since then. He has lots of friends, has travelled the world meeting new people, and is in a loving relationship.

Life is still full of challenges though... Daniel cannot leave the house without counting the number of items of clothing he's wearing; He eats exactly 45 grams of porridge for breakfast (he likes the way the lumpy texture feels in his mouth); and sometimes, when he is doing the washing up and drops a spoon, the loss of control is too much to handle and he has a "meltdown".

For all his social inadequacies he's certainly very good at putting everything down on paper. This book is a very well written chronological recount of the major changes and developments and milestones in his life. It was a real pleasure to read and I can recommend it wholeheartedly.


¹I would highly recommend both authors and, in fact, Oliver Sacks' An Anthropologist On Mars is next on my reading list, currently sitting on my bed, fresh out of the library.