Archive for the ‘Geekism’ Category

Inside Facebook

Sunday, December 4th, 2011

I just caught the last half of a documentary about Facebook, which was fairly decent despite being on the BBC. This link should show you when it’s repeated:

Mark Zuckerberg: Inside FacebookMark Zuckerberg: Inside Facebook TV Schedule

Decent in that I learned something about the various business models that I didn’t know. And the journalist talked about technical details of a subject I understand without being wrong, which is rare. And it wasn’t a hatchet job about the dangers and risks and general awfulness of Facebook and business in general.

In fact the criticism of Facebook was pretty weak. If I “like” a company on Facebook, it can pay to display adverts to my friends that say, “Rob Fisher likes [company]“. This could be construed as using me in an advert without my consent. Or maybe not. The documentary did not seem to have a strong opinion.

And there were academics interviewed who made arguments such as: by keeping in touch with more people, we are having fewer close friendships. Well, anyone can tell for themselves whether that is true, and whether it is a problem.

But the best bits were the bits explaining how Facebook makes money. This included a demonstration of the information available to someone creating an advert. The example was a product for brides, and showed how various filters could be applied (“female”, “engaged”, “interested in beauty”) to see how many people would see the advert. This looked powerful. But not as powerful as the use of Facebook to talk to customers. This segment concentrated on companies’ attempts to get people to leave comments on their pages so that the comments get shown to friends, thereby generating a kind of word-of-mouth advert.

It strikes me that the best way for a company to use Facebook is to let your employees have real conversations, like real people. The documentary didn’t really go there, unfortunately.

Padfone

Sunday, July 17th, 2011

The Asus Padfone is a smartphone that can be docked to a larger screen, turning it into a tablet.

photo of Padfone

And it’s finally a tablet I want, having resisted the first two iPads.

I can carry it in my pocket, but when I’m at home or at work I can dock the phone into the tablet and do everything on a lovely big screen. It’s been argued that you might as well have two devices and sync the data between them, but there are many advantages to a single CPU with multiple form factors:

1) If I’m half way through typing an SMS or Google+ post and decide I want the bigger screen, there’s no interruption. Saving a draft and loading it into a different device would interrupt my train of thought.

2) I only have to manage one set of installed apps. With two devices, I have to install the apps I want on both, and install updates twice too. Similar arguments can be made for device configuration and preferences.

3) I only have to pay for one 3G data connection. These can be had pretty cheaply anyway, but managing two contracts is a pain, and paying twice for what amounts to the same data stream seems pointless. In fact, with this configuration I will use less data — I don’t have two devices both syncing stuff from Evernote, for example.

4) I only have to charge one device. I assume the phone and tablet will charge together, with the phone docked in the tablet. If the phone gets low on juice later in the day, there’s a good chance I can dock it and keep working.

5) It seems like the future. The amount of stuff that can be done on a smartphone-sized device is about to hit some critical level. Already desktop PCs are only needed for high end games and serious number crunching. The PC has become a laptop has become a netbook has become a phone. The only problem is the ergonomics, and a single device with multiple form factors is a good solution.

I have thought of a downside: a phone is somehow more personal than a tablet. If I bought a standalone tablet the chances are my wife would use it a lot at home, too. With the Padfone this is less likely because I wouldn’t be able to do phone stuff while she is doing tablet stuff. This thought is enough to make me consider delaying a planned upgrade of my wife’s phone from Blackberry to Android. Maybe we should become a two Padfone household.

Engadget has more photos.

Sony Vita and VAT

Sunday, June 19th, 2011

The Sony Vita is a new handheld games console, due out by Christmas, that I happen to be irrationally excited about. They have recently announced the pricing for the various worldwide markets.

Explains Andrew House from Sony:

“I think we need to do a better job of explaining issues around sales tax as well. The fact of the matter is the UK price will have 20 per cent VAT included in that, the US price by contrast in LA is without the sales tax included.

If you’re buying it in California you can add another 9 per cent thanks to Arnold Schwarzenegger.”

The 3G version of Vita has an estimated UK price of £279 of which a whopping £46.50 is tax.

I’d like to see more of companies saying this. The worst thing about UK VAT is that it is hidden.

Coffee Price Oscillation

Wednesday, April 6th, 2011

Following on from my analysis of Steve Leighton’s ideas about fair trade, I noticed something else interesting he wrote:

The coffee market has also long been a reactive market; when prices are high, every piece of available land has a coffee plant on it, but when prices are low farms disappear along with the coffee, this in turn raises prices as beans become scarce, resulting in re-planting (you get the picture). For a coffee plant to be suitable for harvesting requires around 4 years from seedling to mature plant. So it’s easy to see why the market fluctuates so much, as there is a four year gap before changes are fully appreciated. The coffee-growing countries have tried and failed to work together, holding back crops to raise prices, but when harvests eventually hit the market a downturn in price always occurs. This leads to the whole boom and bust cycle that has been such a feature of the coffee market for many years.

I’m an engineer, and this kind of oscillation seems all too plausible to me. There is a negative feedback and a delay. It seems inevitable.

But I have questions: in a truly free market, is this inevitable? Can’t speculation such as futures markets damp this effect? If so, why do we still see the oscillation?

More importantly: is there any kind of outsider meddling that can solve the problem of oscillation without introducing even worse problems? (I have a tiny suspicion that the price oscillation may turn out to be a feature, not a bug, but it’s not immediately obvious how.)

Chain Reaction

Monday, March 21st, 2011

Nuclear fission illustrated with ping pong balls and mouse traps.

Coooool.

Navy Physics

Tuesday, March 8th, 2011

I am watching a documentary about the Ark Royal. Some Harriers are practicing strafing with their canon, sending up big plumes of water. One of the ship’s crew says:

You can see the bullets hit before you hear the sound. That’s because the bullets are traveling faster than the speed of sound.

Oh dear. What hope do we have?

Update: When I wrote this, it seemed obvious that the crew member should be comparing the speed of sound to the speed of light and that the speed of the bullets was irrelevant. You see the impact before you hear it because the light travels faster than the sound.

However, Nigel Sedgwick points out that the crew member is probably referring to the sound of the gun, not the sound of the impact. In which case, the relative positions of the observer, the harrier, and the impact point are important.

There are two relevant distances: the distance from the harrier to the impact point (which the bullets must travel), and the distance from the harrier to the observer (which the sound must travel). If they are approximately equal, then the Ark Royal crew member’s comment makes perfect sense.

He has my apologies.

Replicators

Friday, December 24th, 2010

I found, via TV Tropes (be careful following that link), a rather interesting diatribe on the dangers of not respecting science in science fiction.

(Aside: Note that within it, Ken Burnside makes some sensible remarks. That name sounded familiar to me. He is a regular commenter at Armed and Dangerous. This makes me think that a) the bit of the Internet I browse is small and b) I should probably check out some of his games.)

One important consideration is that if you change the laws of physics, there will be unintended consequences:

if you broke the Second Law of Thermodynamics in order to obtain stealth in space, a major unintended consequence is that you simultaneously have allowed perpetual motion machines of the first kind, infinite free energy from nowhere, and all the secondary unintended ripple effects.

However, as an example, replicators are used. And I disagree with the author about the consequences of replicators.

In the sphere of economics, there is the havoc created by the unintended consequences of the Star Trek Replicator.

The producers saw one problem right off the bat, and quickly handwaved a reason which prevented using a replicator to make multiple clones of a person. But they passed over small matter of replicator technology irrevocably causing the collapse of the global economy.

Whoah there! Something that makes everyone inordinately wealthy could “collapse” the economy? At the very least we have some semantic problems here. The way I call it, if everyone has vast to unlimited material wealth, “the global economy” is doing pretty good! An invention that could bring this about would be good for the economy.

The businessman [in George O. Smith's "Pandora's Millions"] says it is too late to suppress the invention, but if the engineers want to prevent it from being a complete and utter disaster, they had better go and invent some substance that cannot be replicated ASAP. Lacking that, there is no way to prevent either currency or cheques from being counterfeited. Counterfeits so good they cannot be distinguished from genuine money.

This supposes that money is represented by some physical commodity that can be replicated. In the case of gold, fair enough. In the case of fiat money represented by paper, then yes, the paper can be perfectly counterfeited and that would be a problem for that currency. But fiat money is just invented by central banks anyway, so there’s no reason you can’t invent some other kind of fiat money.

Allow me to introduce my new currency, RobDollars. The point about money is that is has to be a scarce resource, so for now there will only be three RobDollars in existence. Instead of something easily counterfeit like paper, RobDollars are represented by this table on this web page:

RobDollar1: Bill
RobDollar2: Ted
RobDollar3: Rob

If Bill sells something to Ted for one RobDollar, all he has to do is email me and I’ll update the table by replacing Ted’s name with Bill’s in the RobDollar2 entry. The RobDollars have value because they are scarce, and me and Bill and Ted agree that they have value, and everyone trusts me not to inflate the currency by creating lots more RobDollars. This is no different to how GB Pounds or US Dollars work. But with RobDollars there is no possibility of counterfeiting, and a replicator with be of no use to you for creating money out of thin air, because I am the central authority on who owns each RobDollar.

For a similar approach but with no central authority, see Bitcoin, which uses cryptographic techniques to create unique coins.

So that’s the problem of money easily dispensed with.

With a replicator, everybody can pave their driveway with gold bricks, eat caviar and filet mignon every day, and wallpaper every room in the house with Mona Lisas. Which basically means all these formerly expensive items are now worthless, that is, valueless in the sense of being free.

To which I say, so what? Everyone is richer. This is a good thing. If caviar has value, it’s not because it costs lots of money, it’s because it tastes good. The value of caviar is not reduced because everyone can have it, unless you are serving caviar simply to show off. But there will be other ways to show off, perhaps such as owning the real original Mona Lisa, instead of a perfect replica. Or by throwing a particularly stylish and fashionable party, which you might achieve by hiring a top party organiser.

Of course, if your monetary units are based on gold or something physical, they are now valueless as well. As are any investments, savings, or retirement nest eggs made with such money.

The transition might be uncomfortable, but who cares about losing their retirement nest egg when all their material needs are met for free?

Factories will close sending millions out of work. Who needs the goods manufactured by the factory when all you need is a replicator and a recording of the desired item?

Again, out of work but having all your material needs met for free isn’t really a problem. Idleness might be a problem, but anyone with any amount of imagination should be able to find something to occupy their time. To say that doing a repetetive job in a factory is the only thing keeping large numbers of people happy is to ignore what people who work in factories think about it. A hundred times the wealth without actually having to do the work? Yes please! The kinds of people who are likely to become habitually idle and miserable about it are probably *already* habitually idle and miserable on state welfare anyway.

And it is not as if there will be no work at all to do. Assuming there is no AI, there will be plenty of creative and intelligent work to do. This is exactly the kind of work humans get most enjoyment from, anyway. Manufacturing objects is a waste of human effort as long as there is a possibility of automating it. Working to free people from this is a good thing.

Instead, people can spend their time designing objects. The richest people will be able to afford objects from the best designers, or go and see the best plays, or stay in the finest resorts with the best service. There will be jobs providing all these things.

The stocks and bonds of the companies who own the factories will plummet in value.

Again, so what? No doubt the stocks and bonds of passenger ship companies plummeted when air travel became mainstream. But now relatively poor people can travel the world. This is a good thing.

About the only thing that will still have value will be services. A replicator will not help you if you need a cavity filled or an appendix removed.

Everything will have value, it’s just that everyone will be extremely rich. Services are likely to remain expensive relative to people’s wealth because they will be relatively more scarce.

Some kind of barter system will replace a monetary economy.

I’ve already shown that there is no problem using virtual tokens to keep track of who owns what.

In summary, people will always trade as long as there is a scarcity of something. And replicators, as they appear in Star Trek, only solve the problem of the scarcity of material things. There are plenty of other things that are likely to remain scarce after the invention of replicators, such as services, energy, original artworks and beach-side properties. So there will still be an economy. But yes it will look different and yes you will need to think about what it will look like when you write your science fiction story with replicators in it.

A further interesting question is: what will the world look like with a combination of AI and replicators? The AI can design the object you describe, or generate a movie, novel or computer game according to your preferences. There will be fewer things exclusively the preserve of humans.

Taken to extremes, and assuming people do not value things simply because they were made by real humans instead of AI, and assuming there is enough real-estate (perhaps because of cheap interstellar travel) *then* you might have a situation where nobody trades and there is no economy. The only reason to interact with other humans will be to socialise. But everyone will be living in a paradise of their own design. Sounds good to me, but a long way off.

Kinds of Public WiFi

Thursday, December 23rd, 2010

There is wide-open WiFi. You just connect to the access point and it works.

There is open WiFi. You connect to the access point, open your browser, try to go to your page, agree to the terms and conditions that pop up and then it works.

There is “open” WiFi. You connect to the access point, open your browser, try to go to your page, fill in the stupid form with your name and address and create an “account” with a user name and password that you’ll never use again, and then it works. (WiFi you pay for is always this kind.)

Wide-open and open WiFi is sometimes password-protected. This involves going to the proprietor of the establishment or a staff member and asking for the password, which is usually 1234567890, or something equally guessable, although you will never be able to guess it. If you’re unlucky, you’ll be told it isn’t working, which is probably code for “I don’t know the password”. If you’re really unlucky, you’ll get the response “Wi-what?”

In New York I found many, many wide-open WiFi access points, which are the best kind. This is almost unheard-of in the UK, where presumably everyone is terrified that users will download extreme porn and they’ll be held accountable for it. In the UK, nearly all public WiFi is “open”.

However, the free (until February 28th 2011) WiFi on this Virgin Train is open, which is nice. You just agree to the terms and conditions.

If you connect to WiFi and your non-browser-based Internet connected app isn’t working, it’s because you’re on something other than wide-open. Open your browser, load any page, and complete whatever procedure interrupts you.

A variant of “open” WiFi is catch-22 WiFi, where they send you an email to confirm your email address before activating your account.

Introducing MorseKit

Saturday, November 27th, 2010

Introducing MorseKit, a cross-platform toolkit for learning Morse code. Currently it supports generating random words using a number of letters so you can learn with the Koch method. You type into an answer box, and it tries its best to give you a score at the end.

Since writing a post about amateur radio vs. HomePlug, I got more and more interested in amateur radio. I’ve been teaching myself Morse code, and I had a new Mac, and I didn’t find a tool for it that I like on the Mac.

I was also learning to use Git at the time, and realised that it is easy to write Python/Tkinter programs on the Mac (XCode looks great, but I really don’t want to be coding anything that looks like C in my spare time). Generating the sound of the beeps properly could have been a big project in itself, but an open source program called cwtext came to the rescue for that. Hence MorseKit, hosted on Github.

BBC One HD Formula One

Sunday, November 7th, 2010

BBC One HD launched this weekend. Until now, the BBC had a channel called BBC HD that showed what HD contend the BBC made. Now it has a HD channel that shows everything on BBC One. Programmes that weren’t made in HD are upscaled, meaning the SD picture is shown at HD resolution.

Case in point, the footage of Formula One that broadcasters get from Bernie Ecclestone’s organisation is not in HD, for reasons best known to Bernie. Probably he judges that the benefits to him of upgrading his infrastructure are not worth the costs, and he may be right. It’s still annoying, Formula One would look amazing in HD. Maybe next year.

Anyway, the new BBC One HD channel is still a boon to Formula One fans. Because the picture is much better, even though it is not in HD. Why is this?

Well, the picture that the BBC gets from Bernie is a professional quality un-compressed video feed. In the previous scheme of things, the BBC then compresses this video feed into a low bandwidth MPEG-2 stream of about 2 megabits per second or less.

On BBC One HD, however, the BBC takes the same image and turns it into an MPEG-4 stream of about 6 megabits per second. There are several advantages to this. The first is just that the bitrate is higher. MPEG video compression attempts to reduce the information content of the video in a way that your eyes don’t notice.

This works well, but broadcasters judge that putting out many channels makes them more money than putting out only a few channels. And since bandwidth costs money, they cram as many channels as they can into as little bandwidth as possible. So each channel gets fewer bits per second than it might in an ideal world, and you end up with video that doesn’t look as sharp, or that shows blocky artifacts in parts of the picture that are changing a lot, such as a fast moving racing car.

HD needs more bits because it is HD, so when SD video is broadcast on a HD channel, it gets more bits and a lot of the blockyness and fuzzyness visible on the SD channel is not visible on the HD channel.

Here I have taken pictures of my TV of almost the same frame of video. To compare them, right-click and open each one in a new tab. Then switch between the two tabs. The first is from the SD channel:

The second is from the HD channel.

You should see straight away that the HD image is crisper than the SD one. Here’s a look at a small area of the screen. The top image is from the SD channel and the bottom image is from the HD channel.

In particular the center of the wheel in the SD image has less detail. Having more bits available makes it possible to capture more high frequency detail, like the concentric circles made by the aerodynamic features of a Formula One car.

Here’s another example:

The difference here is caused by the higher resolution of the HD channel. Although the source video is only SD, images are split up into 8×8 pixel blocks before some of the information from each block is thrown away. Because there are more pixels on the HD channel, the 8×8 blocks are smaller in relation to the whole picture, so less information is thrown away, leaving more detail. Notice that the number 5 looks smudged in the SD image, and in the HD image you can see the jagged pixels that make up the diagonal line separating the number from the name.

Here’s a final example that highlights another reason that the HD channel looks better:

Not only is the HD image sharper, the letters U and B are more brightly coloured. Another technique of video compression is to send colour information at a lower resolution than brightness information. This is because the human eye sees colour at a lower resolution than it sees brightness. Because the BBC gets video from Bernie that has not had this technique applied to it, and it then converts SD colour to HD colour, the full colour resolution is preserved.

So whenever you can, watch the HD channel even if the programme is in SD, it still makes a big difference.