Tom Morris

I've been trying to get my head around the SemWeb space - specifically, RDF and it's many abbreviated friends (OWL, GRDDL, SPARQL etc.).

Like microformats, it's very easy to over-estimate the complexity of these things. I've jumped in, and I'm turning out RDF files and have successfully queried one using SPARQL. Next up, I'm going to try and write a schema - hopefully Shelley Powers will help me along (at least her incarnation in Practical RDF).

So, below, I'm providing a Semantic Web reading list. These are links I've found in my search that describe the idea of the Semantic Web and describe it's implementation.

Tim Berners-Lee, Business Model for the Semantic Web.

Tim Berners-Lee, Semantic Web Roadmap.

Dan Brickley, Understanding RDF.

there was a realisation that the world isn't parcelled into distinct metadata communities and that any solution would need to mix and match overlapping data structures defined in multiple application domains

Edd Dumbill, The Semantic Web: A Primer.

Technical peeves aside, the value of the Semantic Web is to solve real problems in communication. First and foremost this means radically improving our ability to find, sort, and classify information: an activity that takes up a large part of our time.

Berners-Lee, James Hendler and Ora Lassila, The Semantic Web (New Scientist).

The Semantic Web is not a separate Web but an extension of the current one, in which information is given well-defined meaning, better enabling computers and people to work in cooperation... For the semantic web to function, computers must have access to structured collections of information and sets of inference rules that they can use to conduct automated reasoning.

Want to see something really cool? Take a look at Semantic MediaWiki, an extension for MediaWiki that allows you to define semantic data inside wiki articles and have that exportable in RDF format.

Comments | TrackBack

Over the last few days, I've switched over to Google Reader. I like NewsRiver, but I find that Google Reader gives me the same kind of functionality, but with a few less annoyances.

For the most part, I use it in almost exactly the same way I used NewsRiver - load it up with feeds, click "All Items" and surf through the river.

I do have a few little things I wish that the Gfolks would add.

Firstly, on the mobile version, when you 'star' something, it should load the next headline. When I'm reading headlines on my phone, I'm doing a simple thing - going through and saying "yea" or "nay". Star means "yea", skip means "nay". Don't show me the same headline again, okay.

Secondly, please, please, please do tagging properly. If you are going to have tags in Google Reader, don't have a controlled vocabulary. Go to del.icio.us, see how they do it, and replicate. The same is true with Gmail. We don't want labels, we want tags. They're different things, you know.

Similarly, the "share" function should tap in to external APIs like del.icio.us. If I click 'share', it should give me a little Ajax pop-in-front window so I can tag the stuff and have it saved automagically in to my del.icio.us.

Release the API. I want to have an offline version of Google Reader. However much we dream about universal availability of access, it's never going to happen. I've said it before and I'll say it again - tunnels. Nice as it would be to use the 'net while in the Channel Tunnel, it's not happening is it?

Have a blog copy-n-post system. Make it so I can replace that silly 'Email' link with a 'Blog' link. And, more importantly, let me set what happens when I click it. I just want some HTML fragments to copy and paste. I don't want to post it to Blogger or through the MetaWeblog API.

Give me my attention. All that attention data are not belong to you.

And make a linkable OPML file available.

Comments | TrackBack

I'm slowly getting in to the Semantic Web stuff. It's a slow process. for one main reason - RDF and SemWeb people have a tendency to multiply acronyms, abbreviations and technical terms like there's no tommorow - hence RDF, OWL, DAML, GRDDL, SPARQL and so on.

I validaed my first RDF file this morning - I wrote it in the super-excellent Oxygen XML editor (I bought a licence for it yesterday). Then I can go on to the W3C website and use the RDF Validator - which not only shows me errors, but also produces a list of all the triples in the document and shows me a PNG file with the data in visual form.

Next I want to figure out how to scale this up using a database. I'm thinking it would be amazingly cool to have a triple-enabled del.icio.us-type system. Imagine - you add a bookmark to your database, and it then prompts you for a whole load of data based on whatever ontologies you've told it about.

It also would show you all the resources in the database which make reference to the bookmark you are adding. Semantic social bookmarking. I'm not sure I've totally got my head around it, but it's worth thinking about.

It's something I've thought about with regard to OPML - basically, have a way to mark a specific bunch of data as belonging to an OPML file in, say, del.icio.us.

If this kind of thing sounds exciting, why don't you hire me to be your R&D department? I can sit deep in the caverns of your company and fiddle about with SemWeb stuff. The more I try to grok this technology, the further I get away from working out how I could turn it in to a business.

I'm writing my RDF in XML, rather than in Turtle or N3 or N-Triples. That's because I'm already used to working with XML, but converting to and from these other formats is something I haven't yet figured out. I know it's possible, but doing it on-the-fly is something I haven't yet worked out.

What would be really cool is if the people who make Oxygen would build in better support for RDF right in to the Oxygen environment. That'd make my day.

Comments | TrackBack

Whoosh. That sound was a whole room full of university administrators gleefully missing the point. Nobody ever said that "Mickey Mouse degrees" weren't economically valuable. That was never the point. What we said was that they weren't appropriate for academia in the way that, say, history or mathematics is.

I have nothing against developing skills relevant to the market. But academia isn't the place for that. I go to university to learn about philosophy. It does a pretty good job of giving me a good overview of the subject and guiding me to read the appropriate things. Meanwhile, I teach myself how to code - mostly because I'd lose it if I had to sit in a classroom learning it.

Academia is there for the provision of academic subjects. That doesn't mean that non-academic subjects shouldn't be taught. Just not in university. Funnily enough, we had this system called polytechnics before. If you wanted to learn academic things, you went to an academic institution - like Oxford or UCL. And if you wanted to learn technical or trade-related things, you went to a polytechnic.

Somebody who knows how to make surfboards provides a valuable skill. But it's not an academic skill - it derives from certain academic disciplines (physics, engineering) as well as certain aesthetic choices. Meanwhile, studying old books is an academic skill. Both are valuable - but in different ways. This new report doesn't seem to actually take much account of this though.

Comments | TrackBack

So, Tony has apologised for Britain's role in slavery. Fantastic.

Quite why, I'm not sure. There's a lot of things Tony Blair has done wrong, but slavery isn't actually one of them, funnily enough.

The idea that people from today should apologise to other people from today for actions performed by people hundreds of years ago is pure superstition. You should be accountable only for your own misdeeds. In fact, it's complete phoniness that a politician can show 'sorrow' for something he's not responsible for, but take no responsibility at all for the things he is responsible for - like the short-sighted mission in Iraq.

The very idea also gives way too much credit to the nation state. The people who were engaged in the slave trade didn't do it for Britain - they did it for money. They were individuals like you and me. Tony Blair can't seek forgiveness on your behalf or mine.

It's the height of arrogance that Tony Blair thinks he can, with a few words, wipe away the evil that was slavery.

The British government should apologise for the slave trade when the German government apologises for the Nazis, the Italians apologise for the Roman Empire and the Israelis apologise for all those barbaric commandments from God in the Old Testament. Yes, all of these things are barbaric. But none of them have anything to do with the current governments or citizens of a country.

Comments | TrackBack

Hmm. Faint glimmers of realisation are starting to appear. I'm not sure how much longer it will take for the television companies to realise that they are totally and utterly fucked. Sorry, but my laptop screen is higher resolution and far more portable than the box in my living room. If you want to do yourself a favour, find yourself an online distribution partner - Akamai, Limelight etc. - and start offering Xvid downloads of the shows for 99p each with no DRM. Actually, offer two versions - the 99p advertising-supported version or the £1.59 ad-free version.

Offer a show subscription option too - so I can drop the RSS feed in to iTunes. And stop worrying about piracy - it's going to happen whatever you do. Trying to fight online piracy is a little bit like chucking rocks at waves or putting drug dealers in prison. It'll cost you far more in lawyers than you'll lose if you come up with a decent business model.

Think about it. Currently, satellite TV starts at £10 a month and goes up to £40. We were paying that £40 a month to Sky for satellite. And it was shit. Too many adverts, way too much crap. I'm only interested in a small number of programmes - and I'd far rather pay £1 an episode to watch them on my schedule than paying £40 a month to watch them on somebody else's schedule.

Television isn't dead yet. But, for me, it's lying on the ground wounded.

Comments | TrackBack

Here's a crazy idea I've had. XPath is extremely useful for screen scraping, right? You can specify your way to a particular path.

The problem is that this path can then change when the site gets redesigned. So, we could build an API and have a particular person, say, do all the transforms. I could find a way of grabbing as much data as possible from MySpace and then provide a clean XML file for you to consume.

Only problem is that I'd probably end up with a nasty e-mail from someone at Fox Interactive - or my site would get a line in their .htaccess.

No, there's gotta be a better way.

That's what I'm now working on. What I eventually hope to have is a way for an individual to make a database of XPath queries for particular websites, describing those in an RDF file. (I'm trying to get my head around RDF and assorted technology, and building something is the best way for me to figure a technology out.)

Then a mashup developer simply has to send out a query using SPARQL to my RDF file, and in there it will contain the latest function name. That way, if the site gets redesigned, all you have to do is go in to the RDF file, change one line, and all the things built on top of it will automatically get updated.

I haven't quite figured out how it's going to work - I've got to read the specs for the Friend of a Friend specification, since I want to base it on that. RDF is good, but it sure is complicated. I'm also having a bugger of a time getting my head around it all. I'll probably order Shelley Powers' book, "Practical RDF" later. That may help.

Comments | TrackBack

Another type of data that I'd love to be able to get access to is course information. Currently this is sitting in a big composite silo called UCAS - the University and College Admissions Service.

Of course, like libraries, universities are engaging in a nice big "boil the ocean" type development project, which I can predict now will be thoroughly fruitless. It's called XCRI and has so far managed to get a small handful of institutions to put up a small trickle of information online in a super-complicated XML format.

It's always easier to write the schema than getting people to go along with it. The more complicated the schema, the more likely this is to be true.

A guy called James Mellor has proposed a far more sensible solution - namely some kind of microformat for course information.

I think this is far more sensible. A modification of hListing for the purpose of semantically marking up university course information would seem far more sensible. The thought that universities would sit down and understand XML schema documents is too much to take.

Instead, if they just slap a few DIVs and SPANs marked up with the relevant class-names on their content, someone like me could write an abstraction layer on top of it with outputs available in whatever formats one needs. Then the fun begins.

I think that these "boil the ocean" schemas have a simple goal in mind - to put off actually doing something. Typical academic move. Spend a few years coming up with the most perfect schema possible, then not implement it because it's too complicated.

For all but about 1% of purposes, getting it out in a rough and dirty format that we have to hack around a bit to get working (like microformats) is fine. It has the significant advantage of being something that might actually be implemented sometime in our lifetimes.

Comments | TrackBack

The other thing I've been thinking about recently is finding a way of doing clever stuff with libraries. I know that many librarians are using OPML to cart library data around - and there are other specifications like MODS XML. Behind library data is a complex protocol called Z39.50, which seems to be a case of an inherently complex system that's been kept alive because nobody can be bothered to think of a better way of doing it (to my mind, SOAP/REST/XML-RPC and XML outputs have gotta be easier than this). There is a library file for using Z39.50 with PHP called YAZ.

Indexdata.com have a guide to Z39.50 servers - check and see if your library is on the list!

David Walker has a project written in ASP.NET (and a port to PHP?) called Shrew.

Some of the folks at the Oxford University are working on making Z39.50 catalogues available via an XML API.

Library catalogues are currently silos. They've got kind of creaky, strange back-doors that make them possible to sneak in to, but they're silos nonetheless. Considering the valuable service that libraries provide, it's such a shame that for whatever reason they are siloed.

Perhaps I've got this wrong - maybe libraries are making their data available in a helpful and useful way for us mashup makers. If that is the case, it'd be one of the best types of corrections and one I would happily be given.

How do you want your library to provide you data? I'd love to be able to have a uniform search facility that would search all the libraries that I'm registered with and sort them by distance from me based on various location rules and settings (ie. if I'm in London, show me London libraries - if I'm at home, show me home libraries). Similarly, I'd love to be able to punch in a list of books that I'm looking for and have an RSS feed tell me when they are available - or when new books get deposited in the library that meet my criteria.

I'd love to have a Greasemonkey plugin that will show me whether the book I'm looking at on Amazon is available from my libraries.

I'd love to have a situation where I could be sitting in my word processor, and I'd be able to just type in the barcode number of the book I'm holding and it'd then build the citation for me.

We are still stuck in about 1986 with regards to the way that academics and students use technology. No doubt, the silo-building technology companies have really helped on that front...

Comments | TrackBack

Kent is wondering whether his Second Life property will make him any money. No, not on it's own. It'll hopefully go up in price so that when you sell your land out, you can make some money. That said, you can use your land to promote a product - like Kosso does with the BlogHUD, and people do by selling all sorts of virtual appendages.

Your property won't make you money, any more than your web hosting account will make you money. You gotta find a way - sell something or promote something from First Life. This is actually very healthy. If you just made money automatically for having property that people visited, then it would reduce the competitive nature of running places in Second Life.

Comments | TrackBack

My Bullshit Detector software is quite difficult to build. I'm searching through text and URLs to produce an index number (kind of like a reverse PageRank) - but it is difficult to detect context. For instance, I consider the words "Deepak Chopra" to be a good sign that "here be bullshit". But how to differentiate between a page that mindlessly repeats his spiritualist twaddle and a page that rips it to shreds? Basically, this is a good case for a lot of testing and iteration.

Currently, I'm covering a wide variety of topics in bullshit detection - pseudoscience (things like creationism and "quantum vibrations"), religion (new age, dark age and revival tent style), alternative medicine (homeopathy and the suchlike), conspiracy theories (chemtrails, Holocaust denial) and the folks who tend to trade in bullshit (Ann Coulter, Prince Charles, Pat Robertson etc.). The other main problem with developing a bullshit detector is that you have to basically don the thickest damn waders you can and climb right in and hunt that bullshit. Currently, I have about 278 terms which I look for (often with variations - I'm using Perl-compatible regex) in order to detect whether the page is bullshit or not. Most of these are names of bullshiters, but also some of the many silly words, organisations and practices they use to foist their rubbish off on us under.

I've built it, and am going to be demoing it tonight to a few friends. If they like it, I'll publish it tomorrow (and, of course, roll their suggestions in to the iterative development process. I'm using microformat-style formatting in order to make it in to a very basic API too. I might also make it so that you can choose to run it in "reporting" mode, which would basically add the results to a database - so I could then output a directory of websites based on their Bullshit Rating.

There is an important task in this - which is to try and shake people in to building web applications that take a stand on things, rather than just trying to remain agnostic. Part of "users and developers party together" is the fact that developers are users too.

Comments | TrackBack

I'm a big fan of Web 3.0. Kent thinks it's silly.

That's the point. Web 3.0 is a silly remark on Web 2.0. At the event I was at today, you could choose some "tags" - little stickers that you apply to yourself to describe what you do. There were tag stickers for Web 1.0, 2.0 and 3.0.

I picked out "semantic", "RSS" and "developer". They're pretty good tags I think. They did have a "penguin" tag.

Comments | TrackBack

A family member sent me this link to an exhibition. Read the description:

Surveillance hopes to explore various facets of the contemporary blurring of boundaries between public space and private life, between what is visible and invisible, between the observer and the observed. Installations excavate layers of photographed and re-photographed images, produced and re-produced identities. Objects that have recently become bearers of fear and the unknown become the subject of art. Overheard (and secretly recorded) conversations become intertwined into an audio piece. As part of a larger discussion around the erosion of privacy during a period of increasing fear, the works in this exhibition have the opportunity to be a part of this public dialogue.

Don't dare give an opinion. Just contribute to "dialogue". This is the same game that theology is doing at the moment - it has learnt never to assert anything because it often gets it wrong, so it now spends it's time in "dialogue" rather than actually worrying about whether what they are dialoguing about is true or not.

Comments | TrackBack

I'm building an application at the moment (which I'm not talking about yet) that is going to have some fairly heavy identity stuff behind it - namely, single sign-on, OpenID and the like.

Today I found, purely by accident, a new scheme for doing page claims called MicroID.

Check it out, it's pretty neat. del.icio.us and last.fm are both using it. I will be too.

Comments | TrackBack

Randy - we've had this conversation before. We had it on Tom Raftery's blog over OPML auto-discovery. And we had it last night in my comments section over OPath. And I'm going to repeat what I said on both occasions: I don't set the standards. I build tools.

OPath works for things with created attributes and without (by specifying a path through the document). I'm not exactly sure how I can make OPath work outside of the OPath domain name. If you want better control of the service, there are two solutions. Firstly, you use XPath yourself to pull data out in whatever way you find preferential.

As I've said, I'm going to release the source code for my script that will be havily annotated. The intention is for a hundred flowers to bloom. With both the source code and with DIY XPath solutions, I am perfectly willing to help people build whatever they want.

I'm not quite sure what Randy's problem is with my use of the URL standard. It's quite a simple URL - you pass two arguments in the URL. The idea is to do it in a way that's less hassle than the old fashioned way. I'm not sure how I should change the URLs in order to conform to standards better. I'll reveal the methods for accessing the URLs in a raw (read, not rewritten) manner in the next day or so.

Thanks for the feedback Randy, but you seem to be confusing me with someone who sets standards. I don't. I build stuff. Setting standards is a thankless task. I'll state my opinion on how a standard should be, but I don't in any way set them.If someone who does set standards uses what I've done to build or amend a standard - then all the better.

Adam has suggested using XML namespaces to have a solution. Some kind of new attribute - like an "id" attribute (podcast.com do this, for instance) - could do the job (albeit, there would be an equal number of problems involved in doing so). If that is the case, then I welcome you all to take your suggestions to the relevant OPML mailing list and persuade Dave to put them in to OPML 2.5 or whatever. That's not my job. My role is building tools. And tool-builders have to live with the fact that their materials aren't always perfect. When I hear people say "Why are you bothering with OPML when it's such a sucky format?", my response is usually "well, I'm not particularly bothered about the suckiness of it - I'm concerned that the perfect formats (RDF etc.) are still, for all intents and purposes, research projects".

If you think you've got a better way of doing OPML ancors than the way I'm doing it, please build it. Or get the idea out there so someone else can build it. I want a solution to the problem. And I know that my solution will get superseded. If it gets superseded by something better, then the aim in it's design has been satisfied.

Adam Curry's initial podcasting client kind of sucked - but the existence of one generally available tool made it so developers perked up and took notice, and start building more tools. That's something that a standards body or schema can never do.

Comments | TrackBack

I've got something really cool to announce, but I haven't got time to prepare the release at the moment.

I've made it so that OPath can work on created dates, as per Dave's suggestion the other day.

The new format is opath.opiumfield.com/*url*/-cc-/*created*

The created needs to be exactly as it is in the OPML. Here is a sample (Grazr).

I'm building some tools to make the process of building OPaths a lot easier. The first of these tools will be released either later today or tomorrow. It's a script for the OPML Editor and it makes directory management in the Editor a lot easier.

Comments | TrackBack

Okay, I've finally cleared out my inbox enough that I can put this script up. It's a modification of Dave's dotOpmlSuite.opmlorg.menuCommands.getOutlineUrl script.

I've put it up in the standard way I put up code - as an outline.

To install it, you simply need to find a place to put it. The "Personal" menu is probably the best place - or maybe the right click menu. Once you've found somewhere, open up the script and copy it to where you want it.

What it allows you to do is select a node in an outline and then run the script on that node. It'll then return the URL for that node through OPath. It's simply piecing together the URL for you - it doesn't need to communicate with the server or anything.

Using this script, the process of making OPath-enabled directories in OPML is significantly easier. I'm also going to write a script to make addressing other people's outlines via OPath a lot easier.

Comments | TrackBack

I've just found a "Web 2.0" service so utterly insane that it makes me want to smash myself over the head with a jam jar just to check that I'm not in a coma.

It's called People2Pray and it is for "managing and sharing prayer requests". Yep, it's applying Web 2.0 buzzword fever - Ajax, tagging, folksonomy, APIs, gradient backgrounds, reflected logos, etc. - to the ancient practice of asking an invisible, undetectable being to break the physical laws of the universe in order to save your Auntie from her debilitating disease. Upon finding that she has recovered, you then thank this invisible, undetectable being for helping her (and try not to think about the doctors and nurses and medicines). In short, it's the biggest "fuck you" to the idea of cause and effect - and the most frequent instance of confirmation bias - that the world has ever seen.

Upon signing up, it seems that the founders of the website have spent too much time on talking to the almighty and not enough time figuring out how to do object-oriented ASP since I get a message informing me: "Object reference not set to an instance of an object". Boom!

I manage eventually to get around this and then logged in. I can now make a prayer request - a simple matter of entering a title and a description. Me being a hell-bound atheist, I stated that my pussy cat has "a real bad case of the clap". I know, I'm horrible.

The rest is all kind of standard Web 2.0 fare - you can connect with friends and 'communities' (there are a number of churches and fellowship groups you can choose from), and you can share your prayer requests with them, or with the wider base of p2p users. Or you can keep them private (feline clap is definitely a 'private' one - as, no doubt, are the "more time to spend on AIM with Congressional pages" style prayers).

Once you've made a prayer request, you can edit, delete or archive it. What's really cool is that when you hit "archive", a box pops up which allows you to choose whether the prayer was answered! "Yes", "No" or "Other/Not Sure". I really hope the administrators of People2Pray make those statistics available to scholars.

I've gotta say, People2Pray is a reasonably well-built site. The JavaScript isn't the most responsive in the world, but it's functional enough. The behaviour of the tag cloud is a bit odd too - I clicked on "postmodern" (because if you are going to go to church, you might as well go somewhere where jeans are acceptable), but nothing appeared. I guess I wasn't praying hard enough - or that God hasn't chosen to reveal himself to such a lowly atheist as myself.

Now, the only problem with the site is that it's based on a crazy premise. I mean, don't they know that prayer doesn't work?

Comments | TrackBack

Adam has an interesting post about coolness and how we distribute both software and the ideas which software sits in a symbiotic relationship with.

I have some thoughts. Firstly, I have bought more software since the Internet made it possible for me to test things out. I shelled out $69 a few months ago on Interarchy - the Macintosh FTP client. That's because I'd been trying it and it made my life a lot, lot easier. Then again, I'm a geek, and I am (a) willing to download software and try it out and (b) not scared of credit card fraud online (how exactly is my card more secure in the hands of a waiter in Real Life or some bored call center operator than it is at PayPal?).

There's no way that I would have been able to go and buy an FTP client in a box. It wouldn't be worth it. What's even cooler about Interarchy was that yesterday it told me on launch that a new version was available. I clicked "upgrade", and it automatically downloaded a new DMG file and opened it up. I just dragged the replacement application in to my Applications folder and restarted Interarchy. That is the most painless upgrade process imaginable.

The last time I bought software in a shop was when I bought Final Fantasy XI. I hate buying offline in about 90% of cases (second hand books are the exception).

Call me a soppy Hegelian, but I don't think that the blogosphere is less open than the old mainstream media was regarding software. Also, the "cool factor" for me significantly decreases the further away from the front end the people are. Back-end developers - sysadmins, DBAs, programmers and developers - have, in my experience, been far more immune to "cool" than the front-end folk. If I'm at a geek event, it's far more fun than if I'm at a designers event. (The reasons for that are obviously debatable, and I expect if I was a designer, I'd, you know, play the whole "cool" card a lot more...)

If Arrington and Scoble's email in-stream is full, this means that they need to delegate. This is exactly what they are doing - TechCrunch has spawned numerous Crunch spinoffs - TechCrunch UK, CrunchGear. Scoble has got involved with PodTech, whose audio and video programmes often show me lots of interesting little nuggets.

And this is an opportunity for a new digg or reddit type site to appear which links to new startups and product releases. I also think that ReviewMe will be a good way for people to get their products out.