Monthly Archives: August 2008

Hiding data in images – Steganography

Either a good way to conceal data, or an under-utilised way of compressing information when sending it alongside images - either way, Steganography is a neat idea.   Sadly, generally used for nefarious purposes.  

For a test, I embedded the text of the U.D. of Human Rights (1772 words) in this image (image courtesy of freephotos) using this site (maxant).  You can decode it using the same site (password 1111).    The text only took 36% of the image, which means that a fair amount more information can be hidden away, with absolutely no difference visible to the human eye:

original

Original

Embedded

Embedded

It’s not too hard to imagine that some really smart uses for this data-carrier method could be integrated into applications that (still) would benefit from better use of bandwidth.

A sphere as a user interface

Flat screen not good enough for you?  Microsoft developsa sphere-shaped one, with uber-multitouch integrated.   Some applications (live omnidirectional camera, so you can spin it 360 degrees) seem to make some sense, others (pong) less so… But I’m all for innovation in HCI.  As is Pat Gelsinger (of Intel, creator of the 486 processor no less), when he says in conversation with David Fearon, PC Plus Magazine:

“[there will be] a resurgence of interest and invention around the user-interface experience… as we enable the UI to become interactive, immersive and intuitive.  We expect that to begin to occur as we move to terrascale computing.”

Though the whole reinvention of the obvious did make me think of the ad that’s showing on TV right now from a car company, asking how things would be if we could design them from scratch again.  They then showed a world with whackily designed lamp posts, water hydrants etc., and eventually their new car.  Thing is, if we designed things from scratch again – guess what, they’d look exactly like they do right now…  In any case, demo of sphere-touchscreen below:

History

As an example of the data that a server can already pick up from a simple visit, have a look at the table below (visit a couple of those sites and come back and see the difference).

I’ve also added a widget to the right of this page which shows some of the other (quite simple) information that’s easily scraped from your visit – such as IP address (which can then locate you geographically, broadly), screen resolution, local time of day and so on.

None of this information should represent a privacy invasion – though you should be aware that it is collect-able (the broader mass market isn’t, generally). 

The great thing is that some of this information is actually very useful at telling a site more about you – and therefore providing you a better and more personal service. Getting the balance of privacy and personalisation is of course a challenge (and one of the most important things we’re working on at at Favy).  Right now, most of the information you see below and to the right is thrown away and not made use of.


Site Visited?

Thanks to gemal for this script.

Cloning me, Cloning you: RFID implants

Ever since Kevin Warwick implanted a chip in his arm in 1998 (the main use of which seemed to be to turn lights on and off in rooms he entered – or at least that’s the bit I remember), RFID chips have been spreading far and wide.

Our most prominent use in the UK has probably been the “biometric” chips in our passports, which yesterday The Times reported as being really very easy to clone.  No suprise there then. 

But a bit concerning if you think of some of the uses suggested for RFID – the most interesting (in terms of learning about people and behaviour) is implanting them, Kevin-Warwick style, under your skin.   A company called Verichip specialises in this, and has implanted them in a Mythbuster, and in this US policeman, who claims his life was saved by the medical data stored on the chip.

Aside from a couple of scare stories linking the chips to heightened cancer risk, the key weakness is the ability to clone and hack them.  Of course, this will probably never be fully solved, and that does mean that if you connect the chip to your credit card details, medical history (although see above) or other area, it could get risky. 

But some benefits are clear

- Medical chipping to identify and feed medical records of a patient quickly, especially at the scene of an accident, etc.

- Ease of entry: the chips are already in use at one night club in Spain.  Taking the chip out of your Oyster card and implanting it in your hand would also mean no more forgetting or fumbling for your card holder on the underground in London.

Tracking: the benefits of knowing where you are start with some typical applications like tracking the whereabouts of prisoners, or kids for their safety, and of course all objects and animals for identification and delivery and so on.  Furthermore, Kevin Warwick’s light switching can be extended to everything around the home of office – lighting, heating, turing your hifi on and off, logging you into to your home PC and so on.  More interstingly, your current location (or at least your location when checked by an RFID scanner) can open up new data streams that you can use to your advantage (or other can misuse to theirs, if so allowed).

Still, I don’t think I’m ready to be chipped quite yet, and if I am, maybe a GPS receiver/transmitter would be better.  If more uncomfortable.

The Oracle

I can remember as far back as 1996 when Larry Ellison started to proselytise the “network computer” – a dumb terminal that provided an interface to processing power that was stored centrally.  it was going to cost $500, which back then seemed a snip for access to that sort of processing power.

Here’s a quote from that article:

Oracle has already acknowledged that the initial price target was ambitious, so it is planning a family of devices, including a high-end machine and an entry-level network computer. The high-end model will reportedly have a keyboard, mouse, flat-panel monochrome or color screen, modem and connection for ATM or ISDN, video conferencing microphone and camera and 4 Mb each of dynamic RAM and flash RAM in a two-pound, laptop-sized box. It will employ a version of the ARM RISC processor from Apple’s Newton MessagePad, giving it the power of an early 486 machine.

Well, you’d expect a bit more than 4Mb of RAM and a monochrome screen for your $500 these days.  But this isn’t meant to be an exercise in nostalgia.  What Larry envisaged is coming true, a little bit more each day.  The apps we can access online are getting increasingly powerful.  And I don’t just mean the things that we couldn’t run on a home PC (eg Google’s index), but apps we’d traditionally run locally.  Of course, it’s all in a bit more of a distributed / cloud-style network of the 21st Century, but still.  Some notable examples are:

  • Photoshop Express.  The app that made me think of this whole note – and the single most impressive Flash browser based product port I’ve seen

Photoshop Express

  • Google Apps – becoming more and more of an Office-competitor every day
  • Apple TV – one day (soon) probably replacing most set-top boxes and linearly broadcast TV (quite a way to go with download speeds and HD though)

There’s a lot also happening in online desktops, but I’m less convinced of those in the short term.

Ironically, as all these apps become better and more powerful, so do the PCs we access them with.  In fact, hardware is becoming so much of a commodity (viz Carphone Warehouse moving into the “free laptop” game) that there isn’t any need any more to go “dumb”.  We can have both – powerful local processing for games and when we want to be standalone, and network-based apps for the always up-to-date versions, and access-from-anywhere-ability.

Going fully network-based is going to take a long while to happen, but with certain apps (TV, for instance – or where the business model means we get great apps for free online), it’ll happen very soon.  Larry, the Oracle, probably feels vindicated.