So how good is Google Mail?

Not great, is Jack Schofield’s verdict. This is a very clear, critical evaluation. He concludes:

“To sum up, Gmail flies in the face of conventional wisdom, which offers two ways to handle email: a slow but simple web-based system, accessible from anywhere, and a faster, more powerful approach based on downloading mail to a PC. At the moment, Gmail comes somewhere between the two. But it is not as simple as a web-based mail service should be, and is not as powerful as a PC-based one can be. If the compromise works for you, it’s a brilliant innovation. If it doesn’t, it could be a terrible mistake.”

Thanks, Jack. I won’t be using it. Could this be the first thing Google got wrong?

Google numbers — why don’t they add up?

Fascinating article by Simpson Garfinkel about Google’s secretiveness. Here’s a quote:

“Farach-Colton was giving a public lecture about his two-year sabbatical working at Google. The number that he was disparaging was in the middle of his PowerPoint slide:

150 million queries/day

The next slide had a few more numbers:

1,000 queries/sec (peak) 10,000+ servers More than 4 tera-ops/sec at daily peak Index: 3 billion Web pages  4 billion total docs 4+ petabytes disk storage

A few people in the audience started to giggle: the Google figures didn’t add up.

I started running the numbers myself. Let’s see: “4 tera-ops/sec” means 4,000 billion operations per second; a top-of-the-line server can do perhaps two billion operations per second, so that translates to perhaps 2,000 servers — not 10,000. Four petabytes is 4×1015 bytes of storage; spread that over 10,000 servers and you’d have 400 gigabytes per server, which again seems wrong, since Farach-Colton had previously said that Google puts two 80-gigabyte hard drives into each server.

And then there is that issue of 150 million queries per day. If the system is handling a peak load of 1,000 queries per second, that translates to a peak rate of 86.4 million queries per day — or perhaps 40 million queries per day if you assume that the system spends only half its time at peak capacity. No matter how you crank the math, Google’s statistics are not self-consistent.

“These numbers are all crazily low,” Farach-Colton continued. “Google always reports much, much lower numbers than are true.”

Whenever somebody from Google puts together a new presentation, he explained, the PR department vets the talk and hacks down the numbers. Originally, he said, the slide with the numbers said that 1,000 queries/sec was the “minimum” rate, not the peak. “We have 10,000-plus servers. That’s plus a lot.”

Just as Google’s search engine comes back instantly and seemingly effortlessly with a response to any query that you throw it, hiding the true difficulty of the task from users, the company also wants its competitors kept in the dark about the difficulty of the problem. After all, if Google publicized how many pages it has indexed and how many computers it has in its data centers around the world, search competitors like Yahoo!, Teoma, and Mooter would know how much capital they had to raise in order to have a hope of displacing the king at the top of the hill.

Google has at times had a hard time keeping its story straight. When vice president of engineering Urs Hoelzle gave a talk about Google’s Linux clusters at the University of Washington in November of 2002, he repeated that figure of 1,000 queries per second — but he said that the measure was made at 2:00 a.m. on December 25, 2001. His point, obvious to everybody in the room, is that even by November 2002, Google was doing a lot more than 1,000 queries per second — just how many more, though, was anybody’s guess.

The facts may be seeping out. Last Thanksgiving, the New York Times reported that Google had crossed the 100,000-server mark. If true, that means Google is operating perhaps the largest grid of computers on the planet. “The simple fact that they can build and operate data centers of that size is astounding,” says Peter Christy, co-founder of the NetsEdge Research Group, a market research and strategy firm in Silicon Valley. Christy, who has worked in the industry for more than 30 years, is astounded by the scale of Google’s systems and the company’s competence in operating them. “I don’t think that there is anyone close.”

It’s this ability to build and operate incredibly dense clusters that is as much as anything else the secret of Google’s success. And the reason, explains Marissa Mayer, the company’s director of consumer Web products, has to do with the way that Google started at Stanford…”.

Accidental customisation

I am an Apple Mac user for a whole range of reasons too tedious to go into here. I like the user interface, for one thing. And I really value the stability one gets from running on a Unix-based system. But there is one thing about the Mac that has always bugged me — Apple’s insistence on a one-button mouse. Not sure of the origins of this, but it seems to have been dogma from the days of the original Lisa design. This evening, though, I couldn’t find my Apple mouse and plugged in to my PowerBook’s USB port a PC optical mouse that happened to be lying around — and guess what! — the right button works just fine, and enables me to all the stuff I used to have to do Ctrl-click for with my Apple rodent. The scroll wheel works fine too. Eureka!

Bluetooth insecurity — an older story than I thought

While putting together the Footnotes for my column in today’s Observer, I suddenly remembered that Bruce Schneier had queried the security of Bluetooth way back in 2000 — when the technology was still mostly a gleam in a consortium’s eye. Here’s what he said then:

“Bluetooth is … an eavesdropper’s dream. Eavesdrop from up to 300 feet away with normal equipment, and probably a lot further if you try. Eavesdrop on the CRT and a lot more. Listen as a computer communicates with a scanner, printer, or wireless LAN. Listen as a keyboard communicates with a computer. (Whose password do you want to capture today?) Is anyone developing a Bluetooth-enabled smart card reader?

What amazes me is the dearth of information about the security of this protocol. I’m sure someone has thought about it, a team designed some security into Bluetooth, and that those designers believe it to be secure. But has anyone reputable examined the protocol? Is the implementation known to be correct? Are there any programming errors? If Bluetooth is secure, it will be the first time ever that a major protocol has been released without any security flaws. I’m not optimistic.

And what about privacy? Bluetooth devices regularly broadcast a unique ID. Can that be used to track someone’s movements?

The stampede towards Bluetooth continues unawares. Expect all sorts of vulnerabilities, patches, workarounds, spin control, and the like. And treat Bluetooth as a broadcast protocol, because that’s what it is.”

Going phishing

From today’s Guardian:

As scams go, the email was hardly sophisticated. “Hello dear client Barclays Bank,” it warned recipients. “Today our system of safety at night has been cracked!!! It not a joke!!! It is the truth!!! We ask you, in order to prevent problems, to repeat registration of your data. Make it very quickly! Administration Barclays Bank.”

But new research reveals the scale of a fraud which has already cost British banks hundreds of thousand of pounds after customers have been fooled by these apparently crude messages into divulging their online banking passwords to organised criminals.

The survey by the email security consultants MessageLabs will make grim reading for the online banking industry, which now boasts 11 million users. The scams, which first hit Britain last August, may range from the inept to the highly complex but they are starting to hurt.

“Phishing scams are pretty sophisticated, it’s high level social engineering,” said Mark Sunner, chief technical officer at MessageLabs. “They must be working because we are seeing so much of it. All the major banks are saying it’s a problem. Initially they were worried about loss of reputation, now it’s loss of money.”

We need psychiatrists here, folks. For this kind of ‘social engineering’ (nice Orwellian phrase, that) to work, people have to be extraordinarily stupid. Yet they are. That’s also why virus writers increasingly rely on gullible users to open rogue attachments as a way of disseminating malicious code. Why do people fall for this?

Moreover…I first came across the term ‘social engineering’ in Karl Popper’s writings — especially The Open Society and its Enemies. He was attacking ideologies requiring wholesale political reconstruction of society (what he called “utopian social engineering”) because they led inevitably to tyranny and totalitarianism. An open society, in contrast, would be based on step-by-step reform (“piecemeal social engineering”), in which every step could be critically examined and corrected. Seems such a far cry from phishing expeditions.

Bluetooth, schmootooth

I once had a Bluetooth phone — a Sony-Ericsson T68i which I got because my geek friends told me it was the only one that worked with Mac OSX.

Well, it kind-of worked, but was fantastically erratic and unreliable. T-mobile replaced the phone twice, but it got to the point where I couldn’t ever depend on it. And if your mobile is your main phone (as mine is) then that’s just not on. So one day last January I dumped it and went out and bought a cheapo, no-frills Nokia which has worked faultlessly ever since.

My main complaint was with the manufacturer, not the technology. Bluetooth seemed to me to be a good idea in principle. It meant, for example, that I could use the phone (when it worked) to drive presentations on my laptop. And I hate wires, so anything that reduced wiring seemed, a priori, to be A Good Thing.

That was before Adam Laurie went to work on it. He’s the Chief Security Officer at a British company, AL Digital. “Before we deploy any new technology for clients or our own staff”, he told C-Net, “one of my duties is to investigate that technology and ensure it is secure–actually rolling your sleeves up and looking at it, not just taking the manufacturers’ claims at face value. When I did that, I found that it is not secure,” he said.

Laurie found that phones are vulnerable to “bluesnarfing,” in which an attacker exploits a flaw to read, modify and copy a phone’s address book and calendar without leaving any trace of the intrusion. The flaw affects a number of Sony Ericsson, Ericsson and Nokia handsets (including my benighted T68i), but some models–including a handful of Nokia phones–are at greater risk because they invite attack even when in “invisible mode” — i.e. when they are supposed not to be broadcasting their presence. For the grisly details, see the web page he has prepared.

On Wednesday last (April 14) the London Times carried an interesting article by Steve Boggan, who went out on the streets with Adam Laurie and found that Bluesnarfing was indeed as easy as Laurie had claimed. It was also intriguing to see the differences between the two main companies affected. Sony-Ericsson put up a feeble spokesbot who first tried to downplay the problem. Nokia, in contrast, were more forthcoming. When quizzed by C-Net, they acknowledged that some of their phones were vulnerable, but claimed that an attack was only possible if the Bluetooth was in ‘visible’ mode. (Wrong, according to Laurie, for some models.) The Nokia spokesman also volunteered some extraordinary news:

If an attacker had physical access to a 7650 model, a bluesnarf attack would not only be possible, but it would also allow the attacker’s Bluetooth device to “read the data on the attacked device and also send SMS messages and browse the Web via it.”

What does all this mean? Well, it’s worrying, simply because mobile phones are becoming the repository for increasing amounts of personal data. If they are not secure, then there is massive scope for mischief. And the riposte that “one can always switch Bluetooth off” is not as reassuring today as it was a few months ago. The UK law which makes it a criminal offence to use a handheld mobile while driving has led to a massive increase in the use of Bluetooth headsets — which of course require that Bluetooth be switched on!

Photoshopping

If all else fails, incompetent photography can be rescued by Adobe. This picture of the view from the top of the Conor Pass comes from a photograph which didn’t hack it as a straight pic (it was too anaemic, somehow). I expect I didn’t think hard enough about exposure times in the freezing wind up there. But running through a Photoshop filter has made it look more interesting. That beach you can see in the distance is a great place for junior surfers BTW — as the kids will testify.

Postcard humour

The kids and I went to Dingle for the day (we were in Kerry for a short Easter break). Walking down a side-street we saw this:

One of the boys said “That’s a postcard shot, Dad!”. A few minutes later we looked in a shop selling postcards, and sure enough, there was a photograph of a small white dog waiting patiently for his owner. The only difference was that the mutt was outside a pub!

The beauty of idleness…

… is that you have time to take photographs like this!

Taken in Adams’s in Dingle yesterday.

Why mobile phones are annoying

Andrew Monk and colleagues from the University of York have performed a study to assess why it’s so annoying when other people have cellphone conversations in public.

“The researchers staged one-minute conversations in front of unsuspecting commuters who were either riding a train or waiting for a bus. In half the cases, two actors conversed face-to-face while seated next to a potential test participant. In the other half, a single actor talked on a mobile phone while seated next to a potential participant.

Furthermore, the actors conducted half of the conversations at a normal loudness level, whereas the other half were exaggeratedly loud (as measured on a volume meter). The actual content and duration of the conversations were the same in all conditions.

After each test conversation, researchers approached the bystanders and asked them to complete a small survey about the conversation. In other words, while the conversation was taking place, the participants didn’t know that they were part of an experiment, but rather assumed that the conversation was the normal behavior of one or two other commuters.”

[Summary of results from Jakob Nielsen’s newsletter.]