Linux On PLAYSTATION 3

I have finally done it. I mean I have installed Linux on my Sony PLAYSTATION3. This was my third attempt (third time lucky :), as they say). And now I am writing this post on the game console within the Firefox browser, using a wireless USB keyboard and the buil-in WiFi connection.

The first attempt was to install Ubuntu following the http://psubuntu.com/ page. This basically goes down to downloading and burning a PowerPC - based Ubuntu setup ISO CD. Sony has an option in the System Menu to set up a third party operating system on the console and following the prompts I had Ubuntu running on the PS3 in less than an hour. Unfortunately the WiFi connection was not working... I browsed a few forums and soon found out the only option to connect over WiFi was to go for the Yellow Dog Linux from terrasoftsolutions.com. I ordered an install CD for $50 + P&P and run the setup next weekend after it had arrived. The WiFi did not work again, but then I had no time to spare digging for the solution. Almost two months later (as my business activities no longer have to run on nitros), I came back to solve the problem. I soon found out I was not able to login to the system (must have messed with the passwords somehow...) so I went for another (third) reinstall. I takes a little les than an hour to have everything up and running. The setup application works flawlessly (no wonder, its target hardware configuration is very predictable...) and after a reboot I was able to log in to the YDL-powered PS3. During the install process I browsed the YDL help area and the WiFi solution is pretty straightforward - requires to remove the networking adapter and install the wireless one. It is not for the classic Vista clickers, as you have to know your WiFi channel number and your WEP key spelled in HEX (no support for WPA yet... - fortunately for such events I have a second WEP-based WiFi network running at home). But after additional 5 minutes the PS3 connected to the Internet and I was happy to surf the Web using Firefox.

Firefox seems becomes more and more THE universal (web) work environment. It really does not matter what hardware / OS combination you have. As long as there is Firefox version matching your environment, most of the applications will run. GMail started without a glitch. Google Reader was next to pass the test (although I could not make the Gears-based offline version run - Gears Firefox extension was reported as installed, but the option to go offline was still not there - anyway I left it for the time being, as it is rather unlikely I would be using the PS3 offline). Third test was Google Documents, I played with a spreadsheet and everything worked as expected. This just proves next time you think of writing a desktop application, to broaden your reach, you should target Firefox instead of any processor/OS specific combination. The Google Spreadsheet looked and worked identical on the PS3 as it does on my x86 Windows laptop.

There are still some drawbacks. Bluetooth intreface is non existant within the YDL environment (I shopped for a PS3 Bluetooth keyboard just to realize I need to have a standard USB one to run Linux...). Simple services as YouTube need complex installation of various plugins. System startup time is long... But I think Sony should strongly embrace the Linux strategy for PS3, making it easy for ordinary people to deal with, not just hobbyists and passionate geeks willing to dig the Internet for support information. There are many people who prefer to stay away from computers, but would buy a game console or a set-top box to be able to use simple Google applications like email, news, reader or YouTube. The other option for Sony would be to drop its own web browser (as it is not so very much compatible with current Web standards) and invite Firefox to the native XMB console user interface. This simple move itself could easily increase the number of PS3 boxes sold.

Where Is Content?

The Wall Street Journal reported on Tuesday Google would launch a search service for mobile content. Content - based services are in their infancy now. The reason is - it is still extremely difficult to find content we really like. Take music. You have to know the name of an artist or a song title. Then you go to google.com and search for it. Or go directly to Amazon or iTunes for a purchase. This is the traditional way. Boring... Music is about emotions. How often we hear something on the radio and scream "I want it now". Artist? Title? Don't know and don't want to wait - I like it, give it to me.

For a long time Pandora has been the only service close to the above mentioned scenario: I like it, give me more like this, now. But with all due respect, Pandora is hand - made in fact. There are people who listen to songs, classify them into categories - no magic, just hard work. Google announcement does not bring us any closer to a content search we would love to have. At least not in this version, as they seem to be focused on being more of a content broker for now, not really enabling us with a new way to search and navigate content. On the other hand Google's move is welcome, as while computer is not the best device to look for content, the mobile phone is. Content searches are unexpected and impulsive in nature. And a cellphone is far more convenient to reach for most of the time. You probably won't have a chance to use computer - based music search while driving a car or watching movie.

A cellphone is not only a natural device to look for music. It is also a natural device to purchase and consume music - in form of a song being played over the headphones while commuting to work or in form of a ring tone or in form of a ringback tone. Actually the lack of a cellphone - based music search service is stopping the potential explosion of mobile content business. And if you think it has already exploded - think twice. How easy it is to find a good ringtone? Not the one from the preloaded gallery on your mobile. Your personal ringtone, expressing yourself. I guess it takes a good hour and some 1% people or less actually do it. Same with ringback tones. Of course you can go to your self service web application hosted by your carrier and select it from the menu they serve you. But how limiting this is! Some carriers allow you to upload your own music clip to be used as a ringback tone. But again - saying 1% of us does that would be an exaggeration. I would love to be able to set a theme from Pirates of the Caribbean as my ringback tone, when watching the movie, not when I come back home and turn my computer on. Difficult? C'mon... It's code should be printed on a cinema ticket as a 2D Data Matrix code to be scanned using the phone's camera.

There is no doubt ubiquitous search and access to content, especially music, from a cellphone would trigger a massive shift in usage patterns. If backed by the Labels, it should set them free from worries of falling CD sales. Ringtones and ringback tones market itself can easy grow by a factor of ten, if coupled with a good, easy and accurate content search and distribution engine. The task is not easy to fulfill, but the potential is rewarding for all parties. A couple of billion people will spend a dollar or two a month on content gadgets. Why not let them do that?

Security Mishaps

I really enjoy watching various security - related threads on the Internet. It is just amazing how completely blind and insecure we were running our machines back in 1995. Sounds almost like the Flower - Power times. Lots of things have changed since then, but people (both lame users and experienced system designers) are still being caught by hackers and intruders. Here is the recent favorite collection of mine:

• New attack cracks WEP in record time. WEP is the old wireless network security scheme. Now it takes two minutes using an ordinary laptop to break it.
• Malware will thrive despite Vista efforts. Mark Russinovich explains why even the latest security measures won't stop malware programs from sniffing your data.
• Satnav hacking made simple. This story describes how easy it is to inject spoofed information into the real time car navigation systems we all rely on.
• How credit card data went out wireless door. This is a good one. Actually one of the most successful high profile credit card data thefts. 45 million credit card numbers in 18 months, probably worth more than a billion dollars and hackers never caught.
• Crook reprograms an ATM to think $20s are #1s. This story shows how so called layer-8 security (ie humans) is the easiest to break. The guy entered a master password into an ATM machine made by Triton. The password ("123456", no kidding) was printed in a manual published on the web and somebody forgot to change it...

There are hundreds or even thousands similar stories. The take-out is common and simple. Understand how the things you are using work. And think. Do not let your password out when a stranger offers you a bar of chocolate. Actually one of my recent hack ideas is to leave an USB stick somewhere at the office. The chances somebody will pick it up and plug in are close to 100%. And then you can have a Trojan application there being launched as soon as somebody clicks on it. No operating system on Earth will prevent a user from doing that, only the common sense will save you...

Voice: The Killer Mobile Application

The world is on hunt for the next killer application for mobile phones and mobile networks. Last week the US went crazy over the iPhone. We talk about mobile Web, mobile video, mobile location - based services, instant text messaging, mobile email, social networks, new generation of mobile games. But let us hold on for a second. What about voice?

What do I mean by "voice"? Try to explain... I talk to the handset and the other party can hear me. And she talks to her handset and I can hear her. The method and apparatus patented by Alexander Graham Bell in 1876, after changing the world forever, is still in use today. The howstuffworks.com site explains in detail how this happens. Important here to note is the quality of the voice conversation. The wired telephones of the past century established a tradeoff between call quality and network economics. To allow more simultaneous calls on the network, the audio bandwidth is limited to about 3,000 hertz. All of the frequencies in your voice below 400 hertz and above 3,400 hertz are cut off. Again, the howstuffworks.com page demonstrates the frequency clipping effect. And it is really hard to believe, despite all the technology advances, we have been stuck at this point for more than 100 years now. In modern vocabulary this is referred to as narrowband voice coding, the technology responsible for transmitting our voice over the telephone network.

Actually thing have recently even got worse, thanks to finance departments and stock exchange investors, forcing the network operators to cut down on OPEX (Operational Expenditures). The OPEX savings are done by using immature VoIP (Voice Over IP) backbone networks and interconnects, often resulting in completely impossible to understand conversations between two parties (especially on long distance / international calls). I have this everyday, when my business partners call me from abroad and I can hardly understand what they say. Often I hang up and call them back (the return call is routed over different path) trying to catch a better connection. Is this the state of the art in voice communications AD 2007? Give me a break...

More and more frequently I use Skype, whenever I'm within a broadband Internet coverage. Not because Skype is free, but because it offers better call quality (hear it, Mr Carrier?). Instead of narrowband, Skype makes use of wideband codecs, whenever possible. The resulting call quality, while far from high fidelity, is far superior. Some mobile carriers are looking into wideband codecs (like T-Mobile with Ericsson). What is funny they claim it a "voice quality hereherto unknown". Funny, because AMR Wideband (the technology they use) transmits only half the bandwidth of a 50-years old FM radio we use everyday. Anyway this is the move in the right direction, but we are still far from that point, and true HiFi calls and voice conferencing for years is going to be available only over the Internet.

I have no idea why this issue has been forgotten for so long, especially as - to quote Greg Papadopoulos, the CTO of Sun Microsystems, once you have taken part of a true high fidelity voice teleconference, you never want to go back.

Cash, Checks, Credit cards,... Telephones

Recently I have been asked what would come after credit cards. First second I was a little shocked with the question, especially as I was just going through a wallet with several of them in it. Why would we need something else? Aren't credit cards fine? To be honest I do not miss any functionality here - the cards just work. I can pay almost anywhere in the world with a piece of plastic, linked direct or indirect to the bank account. Security? Well may be this is an issue, but if it was a real issue, we would have done something about it already.

So I was silent for a few seconds after the question was asked trying to figure out the answer. And then I looked at my mobile phone. Yessss! It is so simple. It will be my next credit card! Then I even reminded myself I blogged on the subject half a year ago :) But it is worth to bring back the conclusions. Mobile phone, evolving towards a personal teleputer, is really going to be the only thing we carry with us. As it becomes a universal (voice and video and text) communication device, a camera, a media player, there are more and more reasons to carry the phone with you. There is a paragraph on using a phone as a method of authentication in the post I refer to above. So It won't be long when you start a car with a push of a button. The car will check for the presence of the phone over some Near Field Communication technology and the immobilizer will let the engine start when it successfully authenticates the owner via the owner's mobile phone. So the keys will be another thing of a past, to be left at home.

Credit cards will eventually disappear the same way as car keys. A personal mobile phone will take over the authentication function. The fact is, simplifying things a bit, authentication is the only function of a credit card today. The card does not pay a bill. It is your bank who pays. With a card you authenticate yourself to the bank and with a signature you authorize the bank to transfer the specified amount of money to the seller. So if a phone can authenticate you to the car, there is nothing to stop it from being used to authenticate you to the bank. A phone also has two important attributes credit cards miss - a display and a keypad. So there can be a dialog. You wave the phone over the cash register, and at this very moment you are authenticated to the bank. To improve security the bank may ask you to enter a PIN on your phone's keypad (your personal keypad). Then the amount you are about to pay shows on your phone's screen with a prompt to confirm the charge. There are a number of ways to improve this scenario later. Like biometrics - a touch of a finger instead of a PIN. But even the basic one is going to make the credit cards obsolete. Why carry car keys with you, why carry a wallet, when a phone is just what you need?

In parallel, banks are experimenting with full mobile banking over the phone. Not just payments, but full home banking - account management, fund transfers... There is a good article on the subject on CNBC website: Banks Go Mobile.

Just wonder what Alexander Graham Bell is thinking of the way his invention develops.... And AT&T 100 years ago probably did not envision itself being a bank... Well it is hard to predict, especially the future :)