kief.com

Workmate Ben (or "Bean", as he calls himself), provides us with this link to a kiwi christmas story. "They came in, said 'Merry Christmas' and then helped themselves."

Google alerts has just informed me of the existing of "Kief Princess", who is a champion breed mare in Ireland according to this article. Too bad Good thing she's not a racehorse, or I'd be betting on her in every race.

I would really like to build or buy a box that lets me easily watch files I download with bittorrent on my TV. There are a lot of interesting new media appliances for the TV these days, including:

• PVRs, which are hard drive-based TV recorders with nifty features for automatically scheduling shows to be recorded,
• XBox and XBox 360, which are game consoles that can buy and update games through an online service
• Windows Media Center, which is basically a PC running a special version of Windows XP designed for PVR and other functions

I don't have any of these things, but I'd quite like something that does at least the PVR stuff, but also lets me run bittorrent.

If you don't know what bittorrent is, it's basically a network protocol for downloading files that is well suited to transferring large, popular files. Without going into the technical details (which, as a geek, I find fascinating), bittorrent turns the usual dynamics of downloading files on their head - the more people that are downloading a file, the faster it will be for you to download.

This makes it an ideal and increasingly popular way to download software, television shows, movies, etc, which tend to be large files. Serving media files on the Internet requires a lot of bandwidth, which gets very expensive for popular files, but with bittorrent it's much more efficient, cheap, and faster, so better for everyone involved. Of course much of the content people download using bittorrent is of dubious legality.

Personally, I like to download TV shows that aren't available in the UK using bittorrent. The two big things I like are that I can watch it when I want, and without commercials. The big thing I don't like is that I mostly have to watch these on my PC instead of my TV. It's possible to hook my TV up to my PC, but not convenient, since they're in different rooms, and the PC doesn't have a remote control.

A PVR like Sky+ would probably give me the two things I like about bittorrent, and solve the thing I don't. But there is another advantage bittorrent has over a PVR, and that's the Internet. A PVR still only lets me watch whatever is made available by my cable company, satellite service, or local broadcasters. A torrent box connected to my broadband Internet connection and my TV would give me access to any content people put on the Net.

This would do for TV what VOIP is doing for phone service, and what IPTV doesn't even try to do) - wipe out geographical limitations. Producers and/or distributors of television content could make their stuff available on the net via bittorrent, either charging a fee or, for promotional or amateur content, for free, and reach viewers everywhere. My wife and I could watch TV shows from Turkey, local stations in the US, and independent producers, no matter where we live. That would rule.

The box is actually trival to make. It's a hardware device that can connect to a TV and a broadband connection (a plain ethernet jack or wifi would work for me), play media files (XviD), and run a bittorrent client (ideally Azureus), and a UI to make it easy to browse and select files to download. Adding 1 or 2 TV tuners would give you a PVR as well, and a DVD drive (preferably a writer) replaces your DVD player.

The quickest way to get a box that does this, and more, would be to get a Windows Media Center device. This is basically a PC shaped and designed like a home entertainment component, running a special version of Microsoft Windows. It would be a simple matter of installing Java and Azureus, and ideally writing a decent UI (or perhaps WMC plugins?) to browse and select torrent files to download. Alternately, the same hardware could run Linux and a free PVR software package like MythTV.

The drawback of this is expense. Most of these boxes seem to cost around £1,000. I'm sure a box could be sold for around £300-400 that could handle this, maybe not quite as full powered and full featured as the media center PC's, but enough for downloading and watching TV. The new XBox 360 games console actually looks to have the right hardware. I think the only thing keeping the 360 from being able to do this easily is that its Internet connection seems to be restricted to connecting to Microsoft's online service. I'm sure there are hacks for this or, at the worst, Linux will inevitably be ported to the 360 the way it has for the original Xbox.

In fact, I could probably get a workaround for what I want by installing Windows Media Center on my normal desktop PC, and buying the XBox 360 and connecting it to my TV. The 360 apparently works with Media Center as a "Media Extender", so it can play music and video on the Media Center PC's hard drive. I'm not sure how easy it would be to install a program on the XBox that let me browse and choose bittorrent files from the X-Box though, and it is overall an awkward solution.

In any case, the hardest part of all of this is the content. Most of the content out there right now is illegal, so nobody other than a hobbyist like me is going to make a bittorrent box. The key to making it popular will be on the content side, which is an interesting business challenge. I've got more on that subject for a later post.

If anyone knows of a box that costs around £300 and can run Linux (or Windows Media Center), has a hard drive, TV connectors, and ethernet or wifi, let me know. The closest I've found in price is the Topfield (also see the UK forum), which is a sweet-sounding PVR but has no network interface. Shuttle makes many different small form-factor PC's, but these are pretty powerful and more expensive than I'm looking for.

My friend Iain is running 10k this weekend for a charity that's helping a friend of his who has cancer. Consider pledging a bit of cash for this good cause.

I have to say though, that picture of Iain running - dude, what's with your arms, they scare me! We need to get some protein into that boy.

I want one of these: The GoodFather.

Adam Bosworth of Google has taken a look at how data is being used across the web, that is, how applications have evolved through popular usage patterns, rather than from a Computer Science view of how things ought to work, to draw lessons for application architecture. His article in ACM Queue called "Lessons from the Web" is oriented towards lessons for database design, but it's interesting that it ties in with many of the so-called Web 2.0 concepts, and I think is worth looking at and thinking about for anybody designing online applications and services.

If you're a Java programmer and interested in adding sexy Ajax goodness to your apps, a good starting point is Matt Raible's presentations, which he has made available for download for those of us who can't actually go see him talk.

update: The pdf presentation I downloaded actually doesn't seem that useful on its own, it doesn't contain as much information as I'd hoped. Most of the slides are just Matt's photos, that I assume illustrate personal ancecdotes that he uses to make points about the topic, but are kind of baffling on their own. I think the equinox-ajax.zip file probably contains code examples, but in the past when I've downloaded Matt's appfuse framework starter kit I couldn't really get into it, and I believe equinox is a version of that. Bram Smeets intro to using DWR may be a better start.

While I'm on the subject of former Syzygians, it seems that Irakli West is starting up a new venture, TRND, The Real Network Dialog (all in Deutsch, unfortunately for us non-German speakers). He describes it as a Web 2.0 based word of mouth marketing system, similar to bzzagent. He's also got one of Germany's top blogs, about Firefighting. Cool stuff, or at least I assume it is.

Fellow ex-Syzygian Iain Tait has been blogging for a month now. Iain's a smart guy who's a Creative Director at Poke, and it's fun for me to read what an old friend thinks about things going on today. He's mainly posting on technical/business trends, things like Web 2.0, and he's got a keen eye for quirky new gadgets and toys. So far Iain has been keeping up a good blogging pace, with frequent posts that are tight and to the point.

Brief posts are really the classic weblog style, and obviously makes it easier to keep the posts flowing. Personally I have a very difficult time writing short posts. Even when I start with a small, simple concept I want to post about, I tend to ramble on. Like starting a post about a friend's new blog, and carrying it on into a discussion of weblog posting styles. The result is I post a long, incoherent screed or two every few months.

Gotta work on that.

Update: Iain protests that he's not a Design Director, pointing to his blog's design as evidence.

London's mobile industry now has a monthly get-together, thanks to Mobile Monday London, aka momo london. I went to the first meeting Monday evening, which I enjoyed.

A show of hands showed the turnout to be roughly half and half techie/biz-marketing, with a fair chunk of designers (maybe 10%). Speakers were encouraged not to make a product pitch, instead there was a theme ("bridging the digital and physical worlds"), and the speakers oriented their talks towards their experiences and learnings that relate to the theme. Naturally they did mostly talk about their own products and used bits of what were obviously their standard presentations, but I think they did a decent job of not just coming in and pitching us.

There was a good bit of general discussion between speakers and audience. A common concern of the group that emerged was the difficulties of getting customers "into" a mobile product, in terms of the way platform incompatibilities, awkwardness of user experience, and legal and other obstacles that make barriers for users to register with and start using a mobile product and service. Since this is something that affects and requires all three of the types of mobile animals that came to this event to work together to solve, Mobile Monday has the potential to be quite interesting.

The organizer was Dan Appelquist, a technical manager of some sort for Vodafone. Russ Beattie also came in from Silicon Valley to attend. I was chuffed to meet him, since we've corresponded by blog years ago, and both still keep up with each others' blog. He prodded me to post more to my site, so here ya go Russ! I also met some other people, but I'm shit at remembering names, which is why I'm not a sales guy.

There were three speakers at the event, so I'll give a capsule review of each.

Dennis Hettema from OP3/Shotcode

Shotcode is one of those schemes where a visual code is put on a print advertisement, which users can point their mobile phone camera at to link directly to the advertiser's site. Hettema was the best speaker at the event, he was obviously more experienced at pitching than the others and so was more polished.

I have to say I'm not convinced that this kind of scheme is really likely to succeed. Hettema positioned this as a way to get around some of the limitations of the mobile platform, in particular the awkwardness for a user having to type a URL into their phone when they see an interesting advertisement. In reality though, this isn't how mobile advertising works, at least in the UK. Ads for mobile services ask people to send a text message with a simple keyword to a shortcode, a 5 digit phone number. When they do this, the advertiser can send a WAP push message, which is essentially a text message with an automatic link to whatever they like. This is a very widely used technique, and works on TV and other media where a visual code like shotcode would be difficult to use. Plus it doesn't need special software to be installed.

It was notable that when Hettema gave his stories of shotcode campaigns, talking about how successful they were, he didn't mention the total number of people that used it, which he would have if they'd had impressive results. Instead he mentioned how many times each user who did try it used the service. The one time he did mention how many people used it was a case where they handed out phones with their software preinstalled at a convention so exhibitors could scan shotcodes on attendees' badges to get their contact details.

Also, the speaker from the BBC gave figures for a campaign they did which used a shotcode-like scheme along with other ways for users to use their service, and the shotcode-like scheme got spanked.

Richard Jelbert from mTrack/KidsOK

KidsOK is a location-based service for parents to track their kids whereabouts. One tidbit I hadn't known is that there are location service aggregators that you can use instead of going to each operator. However, each operator still has to approve of every location based product, and they have different requirements, so it still sounds very painful. Jelbert is involved in a committee which is trying to come up with a standard Code Of Practice (COP) across operators. But it sounds like even with this, the requirements put up huge barriers to converting users.

KidsOK actually sell a product in retail stores, which is a registration kit. This involves getting a form by post, and
then going through a rigamarole with SMS messages to opt in. Jelbert says of the people who buy the kit and start the registration, 42% don't activate their account, and of those who do, 48% don't complete the opt-in.

The possibility of using a J2ME application to make this process smoother was raised, but Jelbert said this would make approval by operators more difficult. Basically, the standard Code of Practice outlines opt-in procedures using SMS, so if you follow that approval is easier, but doing something unusual means a lot of work to get the operators to sign off.

Location based services sound interesting, but very difficult to do in a way that is going to bring in the user. I think it's probably not worth doing unless you have a service that is killer.

Chris Yanda, BBC

Chris was a very non-marketing guy, but obviously into what he does. His story was about a project called "Coast", where they put up a series of signs at several hiking trails on the coast. These signs offered users several different ways to get information about their location via mobile, e.g. directions to other locations, trivia about the location, etc.

The signs were quite busy, crammed with various ways to access the service. The ways to access the information, and what I could gather about their level of useage, were:

• 4,000 Data code (a barcode type scheme like shotcode)
• 3,700 Audio calls (call up for a recorded message)
• 4,800 Web download (the same recorded messages, but d/l from web page, theoretically to load onto your phone and listen to on the hike)
• 137,000 WAP
• ? Text shortcode was the most popular, but no numbers given.

Interesting stuff!