Tom Morris

Today, I launched an unofficial MySpace API. It's a way of getting data out of MySpace in a variety of XML formats. Three to be specific - a raw XML format, RSS 2.0 and OPML 2.0. The RSS format is there to provide a feed of new comments that have been posted to a MySpace profile.

The API supports gathering data from either a person's MySpace name or from their FriendID. Their name is what comes at the end of the URLs people give out (eg. myspace.com/foo would be 'foo'), while the FriendID is a long ID number. My API takes both, and in the OPML and RSS version, it uses FriendIDs to point to other profiles.

The URL structure is as follows:

xml.opiumfield.com/myspace/username/format

Username can be either the username as it appears in the URL or the FriendID. Format can be either 'xml', 'rss' or 'opml'. The OPML is designed to be Grazed, so don't go importing it in to your aggregator because not much will happen.

I will hopefully be adding new functionality to this API as we go on, but none of the existing stuff should change. It may take a little while for the OPML version to get updated - there's a little bit of disconnect between writing Python and writing XSL. Hopefully, I'll be adding FOAF soon.

If you want to build applications on top of this, please use the raw XML and don't try to reverse engineer the other feeds. The raw XML is there so that you can get as much information as possible.

There is a lot of stuff which gets excluded though - I don't include HTML inside comments, for instance. If you want to get the full dancing baby, flaming skull, MIDI, Web 1.0 effect, you need to go to the person's profile. I just pull text out. I don't get friends other than the "Top" friends. This is not because I don't want to. I'd love to get a person's complete friends list, but it's all hidden away in highly obfuscated, JavaScript heavy code that blows my head to pieces.

If you've got suggestions for data you'd like me to pull out of MySpace, please contact me and I'll try and extract it. To make my life easy, use something like Safari Guide or Firefox XPath Checker and produce me an XPath to the data that you want.

It's all built using Python, BeautifulSoup (which rocks, btw - it seems like the sanest way to do HTML and XML parsing in any language I've seen) and a few XSLT 1.0 stylesheets to turn the raw XML in to RSS, OPML etc. The more I code in Python, the more I like it. I wish my web host would support it better - I'm currently putting it in to a cgi-bin and have some proxy scripts in PHP to handle things.

Comments | TrackBack

I could have told you this years ago, but the Institute of Education at my dear old University of London has published a study of interactive whiteboards which says they're pretty much useless. Which of course they are. I was just finishing school when they started adding interactive whiteboards to classrooms. What they generally meant was that most teachers stopped using the whiteboard because they had to log in to a computer, wait a few minutes, load up an application, wait a couple more minutes and then eventually be able to start writing on the board - rather than just being able to grab a pen and start writing.

Once every few months, they'd get creative and have a fairly rubbish PowerPoint presentation. And once we even had an interactive quiz - we were all given remote controls and had to answer questions on the board. Only some of the remote controls didn't work, which meant that the quiz was more a test of whether the technology worked than whether people knew the answers to the questions.

The other problem with IWBs is that they are often low resolution. A friend of mine had great difficulty reading the board because it was heavily pixellated and was extremely hard to read from a distance. Teachers handwriting generally isn't readable at the best of times, but when it's being dragged through a pixellation process on technology that nobody in the room has been trained to use, it makes it all even more difficult. Worse than this though was the fact that the school would try hard to remove the non-interactive whiteboards from the classrooms, so that teachers would have to start using the technology, even if it wasn't appropriate.

The problem with technology in education is that most of the time, it's implemented extremely poorly. And that implementation is often accompanied by policy decisions at the administration level that lack a trust or wisdom.

At my university, we have a highly-restricted 'walled garden' setup on the machines. There are about four applications that they like - Internet Explorer, Word, Excel and PowerPoint. If you try and use anything other than those, you'll not get very far. Which is fine for 70% of users. But I would rather write my essays in Notepad and save them as ASCII text or XML. No can do. I'd rather use Firefox. No can do. I can't even open up multiple windows in Internet Explorer very easily - the Ctrl + N keyboard shortcut has been disabled.

We've got so many poorly implemented web filters too, even though the students at the college are mature enough not to go surfing for porn on college computers. So many perfectly sane Google queries are filtered because they contain an arbitrary keyword.

The solution to technology in education is simple - listen to teachers, not the government. If teachers start crying out that they want interactive whiteboards, give it to them. Many of them won't - not because they're technophobes or Luddites, but because a lot of the time, they're more trouble than they are worth. As for students, trust them. Punish them when they do wrong, but remember that most of these 'lock down' safety procedures cause more damage than they solve.

Comments | TrackBack

Another good post you may have missed recently - Christian Heilmann thinks that publishers are, thanks to Ajax and Web 2.0, republishing old JavaScript books with the Ajax label, but not actually bringing them up-to-date. I've been battling with bad documentation - I can't stand Python's documentation, and what I can't stand more are books that take about nine chapters to tell me how to do an if...else. I seriously need to sit down and compile a super-handy cheat sheet for languages I'm learning.

I'd love to have a set of very short articles entitled "I Know What I'm Doing, Just Show Me The Diffs". That way, if you already know how to code, but want to know enough Python or JavaScript to start hacking, you could read them. You don't have to tell me the difference between assignment and comparison, just show me the syntax. If anyone knows of good primers that don't waste time holding one's hand, I'd love to know about them. Good documentation really can be a scarcity. We have too many 900 page schema documents and not enough two page summaries.

Some time soon, I really ought to review some of the tech books I own and post up reviews of them. For instance, yesterday I found that Apress' otherwise excellent Pro PHP XML and Web Services contains no explanation of the difference between using file_get_contents() and using libcurl, except that you should use the latter when doing HTTP POST etc. But why? I know the reason, but that kind of thing needs to be explained.

Comments | TrackBack

I've just put out the first episode of a show I recorded with Ian Forrester. You can download the MP3 here or subscribe to the feed. I've got a cool way of doing show notes which will be arriving soon.

Comments | TrackBack

Managed, after about an hour of flailing and cursing, to get a podcast recorded with Ian. We chatted about RDF, microformats and other good stuff. I've got it in MP3, and I'll upload it tomorrow. I've got a pretty swish way of doing the podcast thing. You better wait and see.

Warning - it's long, it's geeky, and it's got the same kind of rough hewn production values you've come to expect from the late, great Gillmor Gang, albeit without the four minutes of waffly commercials at the start of each of the nineteen component parts.

It's all about the attention economy. And gestures. And how links are dead. If that hasn't put you off enough, check back tomorrow and you can listen. Probably.

Comments | TrackBack

Today, I've had a lot of getting things running. I've managed to get an IM and IRC client running on my Palm TX. I've managed to get rdflib running on my hosting account. I've managed to get RDFa running in my brain. And I've managed to get the podcast I've been planning with Ian Forrester for a while ready to go. We now have a blog, a feed, a wiki and some prepared notes for the first episode.

rdflib is a great package, and it's very fast. I'm really liking Python, perhaps enough to shelve the learn Perl project, buy a copy of Mark Pilgrim's book and become a Python addict. I like the fact that modules in Python get compiled in to pyc bytecode. It's like Java, but without the whole use-a-nuclear-bomb-to-kill-a-cockroach thing.

My BarCamp talk is coming along nicely - I wrote a large chunk of it yesterday on my PDA's excellent outliner. Today I've been revising it over and over again. Eventually, I'll shift it over to S5. When I give the talk, I'll also record it as an MP3.

Last night, I went to Geek Dinner and met Molly, Ian and a lot of the usual GeeDee crowd. I also found a few places in Brick Lane to leech wi-fi, which is actually practical with a PDA.

Comments | TrackBack

1. Sell music, not subscriptions. There is only one company that does this properly and it's called Apple. Subscriptions don't work. They're antithetical to the very idea of how you buy music. They don't work for low volume customers. Not everyone is a music addict - and the iTunes pricing enables a person to hear a song they like on the radio, type it in and buy the song. Impulse buys don't happen if you've got to commit to a $8 a month subscription,

2. Sell real MP3s or something that is quite easy to turn in to an MP3. Apple does this reasonably well. Nobody else really does it very well. Apple has the biggest market in MP3 players. Make sure your stuff works on them. That means MP3. MP3 is device neutral, WMA, AAC and the multidue of DRM systems aren't.

3. Once a user has bought a track, make sure that the the user can go back and download extra copies. Hard drives fail, computers get stolen, shit happens. The users are buying intellectual property not download megabytes.

4. Offer podcast clearance rights for commercial music. Allow podcasters to register with your site and pay a chunk of extra cash to have the right to play it on a podcast, The royalty collection agencies are totally irrelevant in the podcast realm, so work around them. The one really valuable bit of PodShow is the Podsafe Music Network. If someone has got a song that cost them $1, let them pay 50 cents more to promote the music.

5. Offer an affiliate programme and build an API. Go and look up Amazon ECS. This is letting people use the data of Amazon's catalogue and sell stuff to their users.

6. Top up the albums. If I buy one or two songs on iTunes, I might want to top up and buy the rest of them. Take the album price, deduct from it the amount someone has paid for tracks from that album and then sell them the rest of the album for that amount. And no bloody 'Album Only' tracks. That's stupid.

7. If you have software, bug your users to backup their music to a DVD-R. Not just their purchased music but all their music. It's highly sensible, which is why I didn't do it.

8. If you are building hardware or software to play MP3s, remember that not all MP3s are music and not all MP3s are three minutes long. The iTunes and iPod experience has this knowledge built in. The Zune doesn't.

9. Cut the parental controls rubbish. Or at least make sure that the explicit version actually is and not vice versa (I've seen a rap album on iTunes with an explicit and clean version - only they've messed up and got the two confused).

10. Give the end user a reason to want to buy music, not a reason to not want to. DRM is the best reason not to buy music.

Comments | TrackBack

You guys know that I'm not overtly cynical about the blogosphere or about tech. I'm cynical about a lot of other things - politics, religion, intellectuals, in fact large chunks of social life.

But I have to be a little cynical about user-generated content. It's not really as important as everyone makes it out to be.

I agree with John Battelle, for instance, when he says that the New York Times doesn't quite get it yet. True. Quite a lot of newspapers don't get it, and quite a lot of broadcast organisations don't get it. But what is it?

It - that is, conversational media, UGC, Web 2.0 - isn't as important as we make it out to be. Yes, we all use it. But if you think that we've got very far, come with me on my commute. How many people do you see reading or producing conversational media? Well, there's me. I'm probably reading sometime on my Palm Pilot or my laptop (when it works) from a blog or writing up a post.

But I am outnumbered many, many times over by the people reading the newspapers. And that's because all of the ways that one views conversational media aren't well suited to being consumed inside a metal box filled with other people.

There are a lot of iPod users on the train, but how many of them are listening to independent music and podcasts? Very few, I'd imagine. iPods are still used primarily for listening to music.

We can dream about conversational media, but we shouldn't get our dreams confused with reality - which is still dominated by newspapers, magazines and "old media".

What we need to do is find ways of getting new media in to the places where old media sits. We need a "print aggregator" - which would pull a few interesting, feature-length articles from sites you like and print them automatically. So if you are getting ready to leave work, you open up your print aggregator, hit print, wait a minute or two and grab the pieces of paper. You've got yourself a little magazine to read but it's made up of interesting things from sources you like.

PDA newsreaders need to improve. Online readers like Google Reader are fine, but they need to synchronise with offline applications so that people can read interesting stuff on their Palms, Treos and PocketPC devices. The same is true for video.

Comments | TrackBack

Paul Levinson makes a now familiar argument about television - claiming that television has entered a real golden age, with more excellent drama series being produced than ever before. See Rome, The Sopranos, The Wire, Da Ali G Show, Battlestar Galactica, Lost, 24, Big Love and Weeds as examples of groundbreaking and risk-taking dramas being produced for (American) television. I agree absolutely with the evidence but not the conclusion. I could list numerous British TV shows which fit the same model.

But Levinson's argument fails to take account of the strength of the anti-TV arguments from critics like Neil Postman who never argued that television isn't entertaining. These shows, among many others, are raising television drama to higher and higher peaks of excellence. I'd even agree that they are more complex and require better preparation to follow.

The important criticism that Postman brought was a fairly simple one - that fields of life which are dealt with on television adopt a new type of context - entertainment. Postman said of television: "it has made entertainment itself the natural format of the representation of all experience... entertainment is the supra-ideology of all discourse on television. No matter what is depicted or from what point of view, the overarching presumption is that it is there for our amusement and pleasure".

The point isn't that television is rubbish, but rather that television isn't the appropriate medium to cover non-entertainment topics. And that topics like politics, science, religion, public policy, war, technology and so on aren't best judged by how entertaining they are. What matters about Iraq is that more than 3,000 Allied troops have died and more than half a million of Iraqis have died. What matters about July the 7th was that a small group of dedicated, religious nutcases got on to Tube trains and killed a whole bunch of people. And yet we pick our prime ministers based on whether they play the guitar, listen to Oasis, ride a bicycle around Hyde Park Corner and how polished their "media message" is.

If you think that the excellence of Lost disproves criticism of television, switch on Fox News and watch Bill O'Reilly for half an hour. Switch on Jerry Falwell's show or read up on Oral Roberts, "prosperity theology" and the fact that televangelists spent two-thirds of their fundraising time telling their flock that if they send cash, they'll get a miracle from God who'll give it back. That sort of television is the reason why heartbreaking stories appear every so often of terminally-ill people following faith healers across thousands of miles and donate thousands of dollars to hucksters in flashy suits.

Television doesn't suck if you exclude all the bits that suck. And the bits that suck tend to be the non-fictional bits. Talking heads after talking heads, idiotic commentators (honestly, watch or listen to a book review show sometime and tell me that a group of random people off the street couldn't do a better job), repetetive news cycles, moral panics, phatically mumbling politicians, personality politics. If you exclude fiction and drama, television really sucks. We forget Postman's lessons at our peril.

Comments | TrackBack

Kosso: "Try to imagine a world without Dave. It certainly wouldn't be as hairy. And i think we'd probably be spending lot more time in our email inboxes rifling through mailling list subscriptions." A big 'Thanks Dave!' from me. Anything to drag me away from my e-mail is good.

The reason I want to thank Dave is more than the technologies he's created and/or passionately advocated for. It's more than the fact that he has, for ten years, posted interesting, informative, witty and biting commentary on his blog every day.

It's because Dave has been an inspiring guy to work with - not an easy person to work with, but an inspiring person. Inspiration is far better than being amicable. If you are amicable to everyone, your software ends up like OpenOffice - with about 9,000 useless features and no clear direction.

Dave's inspiration has taken one key form - the idea that programming isn't hard. Once you get over the fact that programming is as difficult as anything else in life, you start to realise that you can solve your own problems by coding. If something sucks, you build something better.

I'm a user who is on the road to becoming a developer. I know what I want, and I know how to build it. And if I don't know how to build it, I can try and figure it out.

So, Dave, thanks. Without your inspiration, I'd be applying for law school about now. You've saved me from that fiery pit of hell, and instead I'm off to study CS and spend my life building stuff and solving problems rather than making money by not solving problems.

Comments | TrackBack

I managed to get a place for BarCamp London 2 which is being hosted by British Telecom. Hopefully there aren't BT employees reading my blog because if they do, I know I'll be denied entry. If you want to keep track of what's going on, check the wiki or add Ian (username: cubicgarden) to your Twitter. There will be more tickets available later on in the week - there are just under 100 tickets still available, so make sure you are quick on the draw when the next available moment becomes available.

What I really like about BarCamp is that it's a hands-on event and you can actually do stuff. I get fed up of going to events where your only interaction is verbal. We're geeks and we exist to do more than talk. BarCamp is great because people actually do. Last time, Sheila taught me all about XSLT, which was really great. She sat with me, helped me write a stylesheet and tested it, helped me debug it and so on. You don't get that either at a big commercial conference or at a pub or restaurant get-together.

On Monday night, I got chatting with a guy who is a police officer who works in central London. I was telling him about BarCamp and he said it sounded like exactly the sort of conference that more industries, professions and walks of life need. He was explaining that all the conferences he goes to about policing and crime very rarely actually give space to the opinions and ideas of the individual officers who are the folks out arresting troublemakers, dealing with the public, helping homeless people get in to shelters, enforcing ASBOs etc. CopCamp would obviously be very differently organised from the technology-oriented BarCamp. I'm not even sure it would be possible, but it doesn't need to be.

BarCamp is there to raise consciousness to the fact that most conferences aren't very good. My new year's resolution this year is to not pay for any conferences - well, maybe £20 as a top limit (as well as to, you know, stop procrastinating, lose weight, finish my degree, get out to the seaside more often). That's because if you pay for a conference, it tends to be a lot less good than one you don't pay for. This seems highly odd to most people because we've had it pushed in to our heads that expensive equals good. But BarCamp London 2 got 150 signups in 90 minutes. No cost to the attendees beyond their Tube fares (two zone one cash Tube fares eats up half of my conference ticket limit!).

Case in point: Ajax2007 in April. Just saw it on upcoming. Lots of people 'watching' it, a very small number of people 'attending' (and some of those are probably only attending because they are speaking. There's a reason for that. The conference is £1,499 plus VAT. The conference with a masterclass is £2,294 plus VAT. Total price of £2,695.45. To learn about Ajax. I'll plump for the £20 book and if I've got questions, I'll find a JavaScript geek at Pub Standards or Geek Dinner. Tuition fees for a year at university are only three grand a year. Is there anything about Ajax which can't be read about in a decent JavaScript book and subscribing to the Technorati tag feed?

When it's a choice between paying nothing for a two-day conference or £2,695 for a three-day conference, it astounds me that people would pay for the latter. Open source, blogging and the Internet have rendered the Big Conference industry utterly irrelevant. Still, I'm glad they exist. While the big companies are sending their employees off to £2,000 conferences, the small companies are reading weblogs, buying £20 O'Reilly books, going to cheap or free conferences, dreaming big dreams, having enormous amounts of fun and eating dinosaur meat for breakfast.

Comments | TrackBack

I'm currently planning a new format called RollRDF. It's a really simple way of putting blogrolls, podrolls and reading lists together in RDF format. It's a format, not a standard.

The idea is that it should be relatively easy to produce blogroll format that uses Notation3. Obviously, there will be tools to help convert RollRDF format in to OPML so that you can Graze it. The idea is that you will be able to put together an N3 file containing your relationships - that means your friends, the blogs you like, the podcasts you listen to etc. And group them together and do stuff with them.

The RollRDF standard is going to be very simple - it will let you define RSS feeds (and it'll do them automatically when it processes the file if you haven't specified one), and it'll let you make some lists and specify members of that list.

Hopefully, it'll be a practical instance of the Semantic Web (and hopefully, it'll be ready in time for BarCamp).

Comments | TrackBack

I downloaded a demo of the Palm Pilot outliner called Natara Bonsai. It's pretty nice, but doesn't export to OPML.

I'll probably write a script later to turn the output from Bonsai in to OPML. It produces tabbed text. Are there any tools available that turn text/x-outline-tabbed in to text/x-opml?

It's a pretty nice outliner, although the keyboard shortcuts are wacky. What ought to be tab and shift-tab are /+I and /+O (/ is Cmd on a Palm keyboard) - for indent and outdent. Moving up and down is done using /+M and /+V (or just by dragging the node with the stylus).

The other thing that's quite neat is you can set up different 'views' - with checkboxes or numbering. I don't have any of that, I just use it like it's an OPML outline.

The application is $15.95 for the mobile version (more if you want the Windows application, which I don't). I'll definitely be paying that once my demo licence has expired. I'll pay double that if it supported OPML.

I'm using it for brainstorming, note-taking and writing presentations (an idea: OPML or tabbed-text to S5 converter...). I managed to download it and use it to take notes in class yesterday. The notes turned out better than my paper notes did. Combined with the free CardTXT editor for Palm, I really have to say I enjoy writing on the Palm platform.

Comments | TrackBack

I'm working on an essay for college at the moment. I hate using Word - and OpenOffice is just a fairly naff Word clone. I can't seem to get LaTeX to work on my PC. I can't for the life of me figure out how to write DocBook XML (way too complex!).

So I'm now writing my essays as beautiful XHTML. But here's the real killer bit.

I've come up with a semantic format for storing bibliography entries. And I'm going to build some application layer on the top of it. Once I've got some time, I'm going to produce a parser to turn my bibliography entries in to XML, JSON and RDF/XML.

Once I've done that I'm going to explore how to make the process of writing essays easier. Meanwhile, I'm in a great situation at the moment - I've got my essay stored as pure HTML on a USB flash drive.

Firefox on the Mac has a really good print setup. And once you've got your head around hacking little bits of bothersome CSS (hanging indents should be easier), it becomes very easy to just have a template, bash in a bunch of text in to P elements, hit save, hit reload in Firefox.

I'll be releasing the template in the next few days.

Comments | TrackBack

Also known as, don't mistake the appearance for the idea.

Valleywag has an article on widgets entitled "HYPEBUSTING: Against widgets". I found it via Ivan at Snipperoo who - no surprise - disagrees with it.

I think that widgets are a particular incarnation of the web as data, but by no means the best example. Widgets are to the data web as MySpace is to social networking - a good first attempt.

Look at the RAM usage of Apple Dashboard and tell me there's not a more efficient way of displaying this kind of data for the user. A typical Dashboard widget uses 30-50Mb of RAM. What exactly is it doing? Downloading some XML or JSON data from a server and making it look pretty. Desktop widgets suffer because of synchronisation issues. How do I know when this widget was last updated? With an RSS feed in a news reader, it's pretty simple - it's got a date and time stamp attached which I can look at.

Unlike RSS, widgets can't be maintained en masse. While various RSS readers are now introducing attention tracking (how many times do I click on Scripting News posts versus Boing Boing posts?), widgets remain lumps of dumb JavaScript. I can't filter my widgets and say "show me only widgets I've interacted with in the last month". And they don't scale. I know people who are producing widgets for their podcasts as an alternative to RSS. Great. My podcatcher can track three hundred RSS feeds with enclosures quite a lot more efficiently than my widget platform can track three hundred widgets.

I know that if I'm hosting my own blog, I'd rather have a script sitting on my server which would pull in the data, cache it, transform it and serve it out as HTML than a JavaScript sitting on the client side slowing down their machine and being potentially incompatible with their devices (your widgets don't show on my phone, by the way). Client-side scripting has a purpose, but displaying lumps of data isn't one of them, especially when it can be done on the server side. Yes, I know that I use Grazr, but avoiding client-side scripting is good practice wherever possible.

Widgets serve only a very introductory role to the web of data. What they point to is something important - there are Internet applications that need customised interfaces and applications. I'm waiting for someone to build a Twitter client for my Palm Pilot so that it uses my GPRS connection rather than SMS to send and receive messages.

What we need to do is think of the different types of content delivery and then try to imagine data platforms that will work with them. RSS serves a constant stream model very well. Widgets serve a different variation on that - a "grab the latest" model, not a "grab everything" model. If you miss the weather yesterday, you don't want to hear about it today. Imagine other models of content distribution and then build applications to meet them.

The introduction of 'grey' answers in to the application space will be one of those things. Currently, it's a choice between "friend" and "not friend", "subscribe" and "unsubscribe". Instead, we'll be making reference to more nuanced judgments - "this person is interesting but I'm not sure about them", "I like this person, but I'm not wild about their friends", "track this, but only disturb me if something really important happens". Widgets - basically dumb lumps of JavaScript - will need to work out where they are going to stand on the data web.

Widgets aren't revolutionary. They're actually surprisingly non-revolutionary. They are at the very mild end of a shift which could be revolutionary - which is the idea of a data web. I was thinking yesterday of how interesting it would be if all companies made all their data available (obviously with some kind of open trust mechanism in place). The number of utterly useless people ('human resources', 'customer services' etc.) who would lose their jobs and be replaced with Perl scripts is quite unimaginable.

Imagine, for instance, if I wanted to go and buy something. I tap in to a little device what I want to buy and how quickly I want to get it. It then goes and pings the shop databases, gathers back prices, then picks out which ones I can travel to (it would know already that I don't drive plus the schedules of people I know who might be able to pick it up for me), works out the best method of transport to get there, checks that what I'm buying is compatible with devices or processes that I already use, checks whether it would be possible to write it off against tax, makes available money in to a secure trading account to pay for it - following a set of rules about where to take money from and who to borrow it from (if anybody). At each waypoint on my journey, it would tell me the most cost or time efficient way of continuing my journey (the bus isn't running, so get a cab or walk). If I get back home and find it doesn't work, it not only sends out standard complaints to all the relevant people, it helps organise a return trip to the customer service department and manages the financial transaction of refunding the money. Couriers wouldn't try to deliver when nobody is at home.

Think about this from the perspective of health. My medical records are on pieces of paper in an office in my home town. Except my dental records. They're everywhere, because I use the dentist far more often than I go to the doctor (I've had orthodontic treatment and a whole load of other stuff too). My dentist has changed numerous times - one retired, the other one went private, then I was living elsewhere for a year, then I came back and the previous guy had left, and then I found out I needed emergency treatment and I didn't have private insurance so I had to find an NHS dentist who would do the treatment. Each of those has a seperate set of pieces of paper in their office.

There are two ways of putting that in a database - either you put it all in a giant database which you then try to hook up to all health providers in the country. This is big and slow and bureaucratic. Why just within the country? What happens if I need emergency treatment while abroad?

The other way is you put all this data in the hands of the individual on an individual database. Each person would pay a small amount each month for the hosting space for their database, and there would be charities and subsidies for people who can't pay. Whenever a business or organisation or the government wanted to do something, they'd send a request to your database. You could set whatever rules on there you want. If you don't trust your doctor, you are only one click away from denying him access to your medical data. This becomes portable, cheap to provision for (so long as the database could store triples, it's ready to go) and virtually hack-proof. Why? Well, if you set up a national healthcare database, that's one huge honeypot to smash your way into. Once you are in you have the private information of sixty million people. But if you break in to my database, you only have access to my data. Okay, you'd know all the details about what DVDs I've rented and you'd know about the type of braces that I had when I was fourteen. But nobody else would be affected.

The web of data goes much further than widgets. They're an early fork in a very long road.

Comments | TrackBack

Peter has pointed to these statements by David Milliband. Apparently, something inside Mr. Milliband's brain clicked and he is now actually acting rather sensibly. The word "organic" don't mean a thing. I've been drinking lots of chemicals this morning. Quite a lot of dihydrogen monoxide combined with citric acid, glucose, fructose, sugar, ascorbic acid. Pretty scary, huh? Yeah, my pineapple juice does sound scary when you break it down in to it's chemical components. The organic movement are pushing superstitious vitalism on consumers while people outside of the Western world suffer from malnutrition.

Save your money - buy the non-organic food and drop the premium you pay to these phonies in to a charity box to install clean water wells in Africa. If you are buying 'ethically', actually try and figure out whether your ethical action does anything. Organic food does nothing except makes your conscience feel clean - without good reason.

These same holist-vitalist asshats are actively preventing experiments with genetic modification. We already experiment in a not particularly intelligent way with plants - we subject them to irradiation in order to force them to mutate and then keep the beneficial mutations. But scientists aren't allowed to directly modify the genome in order to produce modified species.

Organic food is unethical. I refuse to pay extravagant amounts of money to subsidise farmers for producing a product that adds no extra value. With fair trade, you pay extra to actually help people in need. With organic food, you pay extra to push forward a pseudoscientific agenda that harms the rest of the world by preventing innovation in agriculture. Will my parent's generation (the primary consumers of organic food) get that message? Unfortuantely not. The social pressures of the book club and PTA are too much to deny the holy gospel of organic foodyism.

Produce the food locally, make it taste great, try and benefit the poorest off in society in the process. The 'organic' movement adds nothing to that.

Comments | TrackBack

I've just had a Douglas Coupland moment. I realised that I haven't read Boing Boing for a while, and so I'm surfing through the last week of it in Google Reader. I'm sure there's a bit in a Douglas Coupland novel where a character says that he wishes he could go to sleep for a year so he could wake up and read all the back issues of Time Magazine. Blogs have taken that feeling and compressed it. If you miss a day or so (mixture of a super busy academic schedule, a social schedule and a broken computer) then you get a real headrush when you start to try and climb to the top of the feed mountain.

That's why attention metadata matters, folks.

London doesn't need terrorists - the wind does the job just as well. It took me 90 minutes to get from Kensington High Street to Tottenham Court Road the other day. Google Maps says that this is a 3.7 mile journey by road and should take "about 12 mins" to drive. Not last night, kids. There was wind.

Incidentally, the opening chapters of Cory Doctorow's Eastern Standard Tribe is set on the familiar roads of Kensington through Knightsbridge. It is, of course, totally unrealistic - you could never reach fifty miles an hour, even at night (K-Gore is filled with traffic lights). Based on my time yesterday, that would be 2.6 miles an hour average speed. I was listening to the podcast of and Pluckering (that is, reading on the Plucker ebook reading app) Eastern Standard Tribe on the (delayed) train in to London yesterday. Another Coupland moment.

I found out yesterday that if a commuter train is delayed by more than thirty minutes and it has a drinks trolley on it, they give out free stuff. I got a bottle of orange Tango in the process (I was at the far end of the train - the gin had been nabbed by the time the trolley got to me). It's a bit like a tombola really.

For any Americans and other assorted non-Brits in the audience, a tombola is a curious British practice usually found at school fetés and other fairly ineffectual attempts at raising money and/or community spirit. The basic idea is that in order to raise money for some school building project or the other without doing the rather simple step of writing a cheque people bring in all sorts of old tat that they don't want, stick raffle tickets on it and put the tickets in to a rotating drum which they then spin around. You come up, pay a quid or two and pull a ticket out of the drum. If the ticket's number matches up with any of the prizes on the table, you get the prize. It's like going to a sideshow attraction, only without the ridiculously unsafe rides, plastic ducks or inflatable mallets.

Okay, down to business. The technology I've been getting excited about is quite simple - XForms. Ian Forrester from cubicgarden.com was telling me about XForms the other night and they're mighty cool. Take a read of this article to see how cool they are. I've got some huge great big dreams of where one of my currently under wraps applications involving a plugin architecture. XForms could be a significant part of that - especially since it's all just XML. I could imagine how easy it would be to turn either XML schemas (XSD, DTD, RNG etc.) in to XForms or using RDFS/OWL to generate XForms with not much work. Similarly, WSDL, XForms and some form of stylesheet language for XForms could mean a very interesting scenario - automatic interface building for Web Services.

I'm contacting the brains behind Twitter to ask them if they could make a more comprehensive API available. Why? Well, because I've got an application that I'm trying to build but the data I want isn't available from Twitter. If it were to be made available, we'd be able to do Lots Of Cool Stuff that involves group messaging, mapping and statistics. And RDF of course. Can't forget the RDF.

Oh, and I am sure that this video proves that Microsoft really is fucked up. I'm so glad that the operating systems business is a competition between a group of DRM-addled nutcases and, erm, another group of DRM-addled nutcases that has a very pretty GUI.

I'm hoping that in the next few days I will manage to hack my way through a large part of my workload. And once I've got Zurich Insurance, the Apple Store and whatever intermediaries to talk to one another, my blogging may be a bit more often and a bit less schizophrenic.

Today, I went shopping for a Palm keyboard for my Palm TX. I use my Palm device a lot now, what with my laptop broken. A keyboard seemed like an excellent accessory.

Having done a brief Google search, I decided to plop for the PockeTop Palm keyboard. There didn't seem to be any reason it wouldn't work with my Palm, so I shelled out my money.

I've spent about three hours today trying to get drivers. No drivers on the website. Google doesn't bring up any results. Eventually, I find out that the company who makes them has been sold and the website got bought by some huckster who wants to sell driver updates.

Then I start aggressively trying to find a driver. I go on to the Internet Archive and find a page of old drivers. Unthinkingly, I download one and install it. One hard reset later, I regret that. For those of you who don't know about the Palm Platform - a hard reset is equivalent to a reformat. If you are using Palm OS 5, do not attempt installing PockeTop drivers on there, they will fuck things up. So long as you know that, here's the link.

So, I'm taking it back as soon as possible. I have to accept some blame - shopping on Tottenham Court Road is almost inviting trouble, and trouble is what I have got.

If you are selling a hardware product for the Palm or PDA platform, follow my words: setup your website so that it explains everything I need to know - including the fact that it doesn't work with my device - so that when I'm browsing it on my Palm in the store, I don't buy a product that doesn't work and then have to blog about what a crap experience I've had.

It's a shame. The actual device looks great - I'd really like to be able to type on it. Not providing clear details on a public website about compatibility is evil. Selling drivers to your customers is doubly evil.

I've just posted a review of this useless product to Amazon.

Tags: Pocketop