Archive for the ‘Business’ Category

What everyone seemed to get wrong about the Bitcoin crash

Monday, July 4th, 2011

It was certainly a dramatic story. On 19th June, a matter of weeks after the anonymous crypto-currency Bitcoin began to make waves in the wider world, it experienced a crash that made the 2010 Flash Crash look like a blip. Bitcoin critics, even the normally measured Tyler Cowen, couldn’t resist a bit of self-congratulation. When things seemed to have settled down a few weeks later, the commentators started to ask whether Bitcoin was recovering from the crash.

The thing is, there never was a currency crash. There was a security breach at Mt Gox, one of the largest Bitcoin trading houses, which had dire consequences for their customers. But the journalists who wanted to analyse the impact on the Bitcoin market didn’t get any further than tracking the prices at Mt Gox, the very exchange that had just been cracked, and in the process mistook a bank run for a sovereign default. Limiting their view to this, it looked like the Bitcoin economy was in ruins. Looking beyond the Mt Gox exchange even briefly would have shown the rest of the economy was largely unaffected. Retailers continued retailing, exchanges continued exchanging, and coins that weren’t in your Mt Gox account were as safe as they ever were. If you considered Bitcoin to be a reasonable medium of exchange on the 18th of June, there was no reason to change your mind (though double-checking your encryption and backups wouldn’t be a bad idea).

There seems to be one sensible message to take away from the Mt Gox crash: the cyber-criminals have arrived. If Bitcoin ever was lucky enough to fly below the criminal radar, it certainly no longer is. Optimists will probably say that this moment was inevitable, and may even validate how seriously it’s being taken.

Bitcoin has very real, very interesting economic and usability difficulties that probably mean it will never be a viable currency. Suggesting that the recent security flaws in a single exchange undermine it is just lazy journalism.

PayPal random security checks are ridiculous

Sunday, February 27th, 2011

PayPal recently emailed me to say that my account password had been disabled as a “random” security precaution. To re-enable it, I would have to reconfirm my credit card details and then receive a snail-mail letter in order to verify my address, and of course pick a new password.

Now let’s just think about this. When a PayPal account is compromised by the Ukrainian mafia or whoever, we can assume that they make use of it very quickly. Probably within minutes, certainly in less than a day. Who knows when the victim is going to get suspicious and change their password or notify their bank? Sitting on a compromised account has no upside and potentially a large downside.

In order for a particular random cancellation to be effective, it would have to occur by chance at the exact moment the account was compromised. If it happened beforehand, it would have zero effect (the new password would be compromised rather than the old one). If it happened more than a few hours afterward, the account would already be drained and any protection would be useless. The odds of a particular random check providing any protection are astronomical.

Of course, maybe they’re just lying to me. That would be a whole lot better.

The QWERTY keyboard

Thursday, December 30th, 2010

For years, the QWERTY keyboard has served as an example of a design decision taken for technological reasons that outlived its usefulness. Supposedly, the keyboard layout was chosen so as to slow down typing and prevent mechanical typewriters from jamming, but now we use electronic keyboards we’re artificially limiting our typing speed. Can we finally retire this old metaphor? I can think of several good reasons:

  • There’s no good evidence that QWERTY is substantially slower than DVORAK, indeed the QWERTY layout succeeded in a competitive marketplace against other keyboard layouts
  • It was never true that QWERTY was designed to slow people down anyway; it was designed to reduce the occurrences of subsequent kepresses being nearby in space, not nearby in time, the former being more important to preventing jamming than the latter
  • We have a far better metaphor now in the shape of Twitter

Allow me to explain. Millions of messages a day are now being shared via Twitter. Some people use it to communicate with their family, debate political ideas or get the daily news. Central to the Twitter model is that messages are strictly limited in size, to which many people ascribe its approachability and rapid growth.

But there was never any thought put in to what should be the optimum size for a Twitter message. No studies were done of what the trade-off is between messages long enough for rich communication and short enough to discourage excess verbosity. There were no competing systems. The founders of Twitter simply settled on 140 characters because it was envisaged that Twitter would heavily use the SMS system, and SMS messages are limited to 160 characters (truncating Twitter messages at 140 characters allowed for some metadata to be attached). It’s a technical limitation driving a supposedly human-centric tool.

But it’s even worse than that. The SMS system that set the boundaries for Twitter is itself a holdover from an earlier technologically limited era. SMS messages were originally limited to 128 bytes by the signalling formats used on the networks. Even though this was eventually extended to 140 bytes (the now-familiar 160 7-bit characters) I’m assuming the technological tail was still wagging the ergonomic dog. SMS was envisaged as primarily for traders to send terse stock market tips, not as a replacement for other forms of human contact (fact: you can contact The Samaritans for support with suicidal feelings via SMS; I can’t imagine a worse situation to be trying to repeatedly re-edit your message to fit it into 160 characters).