paul walk’s weblog

Yesterday I went along to Mashed Library UK 2008 in London. Quickly abbreviated to ‘mashlib’, the event was the brain-child of Owen Stephens. Owen did most of the organising, aided by David Flanders who provided the space at BirkBeck college, and our excellent events team at UKOLN. The event was sponsored by UKOLN, using funding from the JISC.

I thought the balance of activities on the day was excellent - a healthy mixture of short presentations, demonstrations and a good amount of hands-on hacking. The group was comprised of commercial vendors (Talis, ExLibris, OCLC), academic-library folk (the majority), a lone representative from the public library world (Paul Bevan for the National Library of Wales), and a few developers from various (mostly JISC-funded) services.

Rob Styles from Talis gave us a demo of the Talis Platform. There is an open API which you can play with - it’s quite impressive. I was very struck by some of the language Rob used in his demo - he talked about dipping, where a result-set from a query (in RSS 1.0 format) is “dipped into” another - with the original data-set accreting more infromation from the second. (Jim Downing and I had an interesting chat about this over lunch, with Jim proposing that we could visualise data-sets as molecules - having a certain shape which allows them to bond with other molecules which have a complementary shape). Rob also talked about mixing in in a smiler vein. The Talis Platform APIs appear to be quite RESTful, with a good deal of passing URLs around rather than result-sets. I plan to have a closer look at this.

Timm-Martin Siewert spoke next about the ExLibris Open Platform. I did get a URL for this but it takes me to a page whcih challenges me for a username and password which I do not have. The Open Platform is , apparently, open to paying customers only. Edward Corrado suggested via a tweet that:

I think they mean open in the sense of the open systems movement of about 20 years ago

Next up was Mark Alcock, standing in for Tim McCormick and representing OCLC, to talk about the WorldCat Developer Network. Mark came armed with a bunch of limited life API keys, so that people could try out some of the WorldCat services. OCLC appear to be offering a spectrum of services, from the commercial pay-for-use variety, to the ‘affiliate’ model - i.e. form a business partnership with us and use our services, to some free services. I’m interested in several of the WorldCat services but am wary of getting too fond of something I cannot, in the end, afford to use. Unfortunately, I did not get time on the day to make use of Mark’s API keys.

I noted that the three vendors represented seem to be spaced evenly along a spectrum of openness, with Talis at the ‘very open’ end of the spectrum, ExLibris at the ‘closed’ end, and OCLC (specifically WorldCat) somewhere in between. I can’t yet see how Talis are going to monetise the completely open model, and I think ExLibris will certainly need to open up somewhat. Perhaps OCLC have hit a sweet-spot of openness? I really don’t know enough about these services in detail, but I noticed some comments from Dorothea Salo which are somewhat critical about the business model behind WorldCat.

Ashley Sanders followed, with a quick description of an Atom (APP) based object store he is developing as part of his work extending the COPAC service. I’m following COPAC developments with interest - I’m very much in favour of the general direction they seem to be taking (I recently blogged about one aspect of this).

Tony Hirst, mashup maestro, gave a tour-de-force demonstration of using Yahoo Pipes and Google Spreadsheets as mashup tools. This went down very well with the technically-minded-but-mostly-not-developers group - especially Yahoo Pipes. I gave a presentation at the Shock of the Social in March 07 where I remarked that the potential of Yahoo Pipes was to do for web development what the spreadsheet did for non-web development before it (Microsoft Excel has been described as the most widely used Integrated Development Environment). Tony showed us how the spreadsheet is certainly relevant in a web-mashup world with his demonstrations of using Google Spreadsheets to mashup data-feeds.

Later on, after lunch, the group got down to some general hackery. On Twitter, Chris Awre (who wasn’t at the event but had been following comments on Twitter) remarked:

Silence from #mashlib08 this afternoon. The mashing must be going well…

And he was right! There was a fair stream of Twitter commentary in the morning - but it dried up as people got absorbed in hacking code and testing interfaces. I saw people exploring the Talis Platform and, in particular, Yahoo Pipes. I expect there will be some blogging about this activity - look out for the official tag:

mashlib08

Andrew McGregor of JISC has already written up his experience of this , as has Jo Alcock - I think these posts describes representative experiences of the event.

Paul Bevan rounded off proceedings with a view from public libraries - the National Library of Wales to be precise. I learned a lot from this presentation about the unique challenges facing the public non-academic sector.

I thoroughly enjoyed the day - kudos to Owen for getting the right balance of people, subjects and activities. There was a ‘buzz’ generated as the day went on which was excellent. I have been to a fair number of ‘hacker’ events where the emphasis is on the tools and the running code - I generally enjoy this kind of thing. But mashlib08 was different - what was really good about this day was that the enthusiasm came from doing stuff with information, more than from the actual development.

I think Tony Hirst deserves a special tip o’ the hat for firing up a real enthusiasm for mashups on the day.

We should definitely do this again!

Technorati Tags: ExLibris, google-spreadsheets, mashlib08, OCLC, Talis, ukoln, Yahoo-pipes

I spent half an hour this morning experimenting with QR barcodes, prompted by Andy Ramsden who is running a small test/survey. I used various iPhone clients to try to decode and make use of three QR codes printed on a sheet of paper. Each of the three codes encoded different information - a URL, a simple string of text, and an SMS message with mobile number respectively.

It transpires that the iPhone does not make a first-class QR decoder. There may be several factors involved here, but the main one seems to be the rather poor camera which often lets the iPhone down. Having tried several (free) clients with mixed - but generally disappointing - results I settled on ‘Barcodes‘ which works rather well, insofar as the iPhone allows it to. One important tip with the iPhone is to take the photograph from a distance of around 18 inches from the QR code - this is counter-intuitive, but it works better within the tolerances of the fixed lense and means that you then have to stretch the image with an iPhone ‘gesture’. Again, this actually worked quite well, but it is a shame that all of this is even necessary. My Nokia-toting friends tell me that it works so well on that platform that it is actually fun, rather than a little chore. Having said that, once the image capture stage is done, Barcodes on the iPhone was actually really good. It interpreted codes correctly, figured out which applications to launch (Safari web browser or SMS client) and was generally well designed. I won’t comment further on the details of the experiment and the results as Andy is going to write this up himself.

So, QR codes - what are they good for? There’s clearly some interest - I mentioned what I was doing on Twitter and got quite a bit of interest. But it’s still rare to come across QR codes in the wild. I see them occasionally on blogs/web-pages but I just don’t much see the point of that (except to allow people like me to experiment). I see QR codes as an interim technology, but a potentially useful one, which bridges the gap between paper-based and digital information. So long as paper documents are an important aspect of our lives (no sign of that paper-less office yet) then this would seem to be potentially useful.

Mia Ridge, who joined the Twitter discussion has also blogged some thought about this - linking to Tony Hirst who mused about embedding links to video clips in QR codes in the margins of paper-based learning materials. Interesting idea? Not entirely convinced, but Mia reckons she would use this.

There seem to be so many factors at work here. If I had a Nokia, with a small screen but quick & direct QR reader, then Tony’s idea would make more sense to me perhaps. With my iPhone, and it’s wonderful big screen and Safari browser but poor QR support, I’d want to read one QR code at the start so I already had the accompanying website for the learning material/course/lesson and be able to navigate around on the device, not on the paper. This is a different model to Tony’s - his is driven by making a direct connection between one section of a paper document and single digital artifact.

Nonetheless - there are plenty of similar opportunities. Imagine walking around a museum - scan a QR code attached to an exhibit, load the URL and get a commentary played on the iPhone without needing to supply/hire those dedicated units some institutions supply to visitors.

The client end of this type of system still has a way to go I guess….

Technorati Tags: qr qr-codes andy-ramsden mia-ridge

It’s Sunday evening….

Brian Kelly recently resurrected the debate about Facebook and its use in an HE context. I know he’s on the road at the moment so I suspect he dipped into his blog post ‘reserve’ for this one ;-). My initial reaction was to smile and move on, but I was caught more by a couple of the comments, from Mike [comment] and Marieke [comment] (both people I know and respect), which have stung me into responding. (Brian is becoming a master at inviting comments of that sort, and his blog has a sufficiently high profile that the comments can invite a response like this one).

There are two messages in the post and in the two comments:

• Facebook can’t be all that wrong because millions of people have accounts in it (”100 million users can’t be wrong”)
• If you say that you don’t like or want to use Facebook, then it is because you are an elitist or a techie or both (and you should “grow up”!)

I have already characterised Facebook as a a walled garden. I don’t feel particularly inclined to advocate the use of Facebook to support activities in HE. I wouldn’t stand in the way of people wanting to access Facebook but the argument which says that university staff should ‘go where the students are’ is often raised but never really backed up - in fact, as Owen Stephens blogged a while ago, there is evidence to the contrary.

Perhaps I am elitist - not for me to say really. I suppose I may be a techie - not sure what definition to check. However, my reasons for not liking Facebook are, I think, reasonable and considered.

As for millions of people using Facebook: well, 2,258,843 copies of the Daily Mail newspaper (which to my eye appears to be a horrible right-wing rag of a newspaper) were sold in August of this year. Hmmm…. more than 2 million you say…. I read the Guardian, but its figures (332,587) just don’t match up…. perhaps I ought to start reading the Daily Mail and recommending it to students?

Mike asks, “if they are [wrong], who cares?” I, for one, hope that our universities do!

This coming Thursday (10/04/2008), UKOLN will be celebrating its 30th anniversary, in an (invitation only) event at the British Library Conference Centre, London. Participating will be current and ex-staff, and a wide variety of people with whom UKOLN has worked or collaborated in some way, both nationally and internationally. In addition to a celebration, in fine UKOLN tradition we will have a series of presentations from senior figures in the Library, Higher Education and Cultural Heritage sectors, offering us their memories and perspectives in a ‘Celebration of the Changing Digital World’. As a relative newcomer to UKOLN, I’m looking forward to meeting some important figures from UKOLN’s past.

I know that at least some of our invited guests maintain a blog, and I hope that some of these will blog about the event. If there are any ex-UKOLN staffers out there who would like to reminisce about the good old days, that would be great. Can I suggest that anyone who blogs, uploads photos from the event to Flickr etc. uses this tag:

ukoln30

That way, we can publish feeds of other people’s memories and thoughts, alongside our own materials which we will be using on the day and making available subsequently. For those of us who like to Twitter, can I suggest that we use:

#ukoln30

Thanks. And if you’re coming, hope to see you there!

Technorati Tags: ukoln, ukoln-anniversary, ukoln30

Some time ago, several of my friends in Facebook installed the ‘MyQuestions’ application: this application allows the user to pose a question and invite answers from their friends. Significantly, in order to answer a question the friend must, in turn, install the application. I was that friend. Since installing the application I haven’t intentionally used it to pose any new questions myself. However, it seems I have posed a question - on installation the application is set to ask a question, any question, in order to be seen by my friends. Anyone looking at my profile will see this question and be invited to answer (after installing the application themselves of course).

So my friends will have seen that I had asked a question I never actually would have asked - in fact the teeth-curlingly banal “What is the most romantic place you’ve ever visited?”. I only discovered this when a newly ‘befriended’ friend obliged me by answering it.

This kind of marketing and propagation of Facebook applications has been described as ‘viral’. Actually, I would borrow a different idea from biology - the Selfish Gene[1]. I suggest that in Facebook we are seeing something slightly new - the wholesale deployment of applications whose over-riding purpose is to be copied and installed elsewhere. Marketing has always been an aspect of commercial software development, but perhaps these applications show a different balance of priorities.

Let’s examine what happened here: An application I installed displayed a message to anyone who viewed my Facebook ‘profile’. Both the application and the underlying Facebook platform colluded to present this message as having come from me. The motivations behind this behaviour are interesting:

• the application is behaving in an entirely selfish way - it just wants to get replicated by encouraging my friends to install it
• Facebook itself benefits from the more applications = more users = more page impressions = more advertising revenue calculation.

What I find disconcerting about Facebook is how we, the users, seem prepared to give up so much control so easily. I notice that some users are starting to use Facebook to serve their personal ‘homepage’ on the web. In the course of this it seems to me that they are compromising in the following ways:

• they have no easy way of knowing how it actually looks to the rest of the world
• applications they install into this area may well be carrying on with their own agenda
• they accept that only Facebook users can access it
• they trust the application developer not to do something naughty on their behalf - asking a very provocative question for example could get more people fired up to respond for which they would need to install the application. Note that I am not suggesting that the developer of this particular application would do, or has done, this - just that the potential is there.

For the curious, the only response to my inadvertently asked question was, “Aaah now that would be telling ;-)”.

The MyQuestions application is developed by Slide. Facebook being what it is, there is no useful way of linking to the deployed application, without requiring the viewer to have a Facebook account.

[1] With apologies to Richard Dawkins.

Update: Mike Ellis, in a comment below, has pointed to this story on TechCrunch, which explains some steps that Facebook are taking to restrict these sorts of nefarious activities in the part of applications running on their platform.

Technorati Tags: Facebook, MyQuestions, selfish application, selfish gene, viral

I recently got spammed invited to participate in Karl Bubyan’s Six Degrees of Separation application in Facebook. This application navigates the ’social graph’ in Facebook, offering a couple of tools to allow the user to test the ‘Six degrees of separation‘ hypothesis. The application and its interface seem quite slick - and it has the now obligatory visualisation (reproduced here).

There’s an irony here. The six degrees of separation idea can only work if there is some barrier to being directly connected to someone else. In the real world, where relationships are subtle, complex and often not immediately apparent, the game of trying to trace the connections and counting the ‘hops’ from one person to another to reach the ‘target’ person can be diverting. On Facebook, at least so far, there is only one way in which users can be connected, and that is through the ‘friendship’ model. In such a simple model, the degrees of separation between any two users are relatively easy to calculate. Although a user can be several degrees of separation removed from another, the barrier to them becoming directly connected is very low - all they need do is become Facebook friends. And how many Facebook users have turned down a request for ‘friendship’….?

So, I suggest that for any actual person in whom you are interested, Facebook presents two degrees of separation:

• infinitely separated (the other person does not have a user account on Facebook at all)
• a separation of zero (you are Facebook friends)

Will these global ’social-software’ tools such as Facebook gradually make the ‘degrees of separation’ notion irrelevant?

Still, this tool is quite interesting in terms of what it allows the user to discover about their relationship to other, named, users, before the user immediately renders it irrelevant by ‘befriending’ them!

And it’s surely interesting that, within the population of Facebook users who have installed this application (4.2 million at the time of writing), the average number of degrees of separation is, wait for it…… 6.12!

Technorati Tags: Facebook, Karl Bubyan, social-software, six-degrees-of-separation, social-graph

I very much enjoyed the UK e-Science All Hands Meeting 2007 last week. Being new to many of the disciplines covered there, I went with an open mind. I learned a bewildering amount, and realised that there are all kinds of opportunities for aligning my professional interests with those of many from the e-science communities.

Some small, specific points:

• Being more used to conferences in the e-Learning an web-development worlds, I was struck by the ratio of women to men. Without having counted in any accurate sense, I estimated there to be an order of magnitude more men than women. Not unexpected, perhaps, in a conference devoted to the sciences…. but worthy of comment I think.
• A frequently occurring theme was the issue of the management of massive data-sets. This community is starting to see that it is going to struggle to cope with the volumes of data it generates. There is a continued and growing interest in data-management, metadata & annotation and compression, as well as much discussion about making such data available for sharing. I was told that the issue of data management has taken over from the issue of providing/using large ‘compute’ resources via the grid - this is now seen as ‘done’ by some people I spoke to.
• Facebook is popular among scientists - I saw wide-spread use of it during the conference. I wonder if e-Science folk are less bothered by the walled-garden issue working, as so many do, with what is effectively ‘closed’ grid technology? Just a speculation…. Apparently, the use of social-networking tools might even be influencing the use of grid-computing, through the way in which virtual organisations are created and managed.
• Java portals seem to be a very popular delivery mechanism for e-Science resources - much more so than currently in e-Learning where they have, to some degree, become out-dated.

Criticisms:

• The conference web site. I can’t even link to a persistent URL for this year’s conference - the best I can link to is this page which will, no doubt, be replaced at some point with information about next year’s meeting. Really disappointing. Update: appending ‘2007′ to the generic URL does seem to work - thanks to Monica who explains this in a comment (below).
• The lack of a tag for the conference - I was forced to invent my own, ‘escience-ahm-2007′, although as no-one else has picked up on it to-date, this was probably in vain. I later discovered that Peter Murray-Rust had done the same, albeit using a different tag.

Update: Interesting to see Andy Powell commenting on the issue of tags for conferences in general.

Technorati Tags: ahm2007, escience-ahm-2007, facebook, grid, peter-murray-rust

Day 3 kicked off with a really impressive keynote on ‘Medical Visualisation between 2D Images’ from Professor Anders Ynnerman. CT scanning is developing to the point where the volume of data returned from a scan of a human body is difficult to manage and process. Various strategies are being developed to reduce the amount of data which must be analysed from a typical scan before the analysis and visualisation processes are invoked. A particular area of development is in supporting autopsies - creating a ‘forensic workflow’. The police in Sweden can deliver a body to Ynnerman’s facility, still in its body-bag, and a scan can be taken and rendered in one minute. We were treated to some fascinating (and occasionally gruesome) images, demonstrating how this technique is ideal for visualising fractures, foreign object (bullet fragments for example) etc. An interesting side-effect of the fidelity of the scans is that bodies are clearly recognisable - the slides we saw had been censored in the traditional way with black rectangles obscuring the eyes of the victims. Other aspects of ongoing work in this area included handling light better in the visualisation process in order to create shadows - we saw how this aided greatly in depth perception, capturing movement over time, as well as some interesting uses of Haptic technology (force feedback devices) to allow surgeons to physically interact with the data, or “touch a beating heart” as Ynnerman put it.

I followed this up with a series of workshop sessions on such topics as static analysis of mobile code, security, metadata and ISO standards.

The image above is of a table decoration at the conference dinner which was an excellent meal - the food at the conference centre has been of a pretty high standard. And Helen, god bless her, kept me up to date with the England V Russia match via SMS.

Technorati Tags: ahm2007, CT scan, escience-ahm-2007, haptic

As an increasingly serious practitioner of David Allen’s Getting Things Done (GTD) system, I immediately appreciated this post on ‘personal unit tests‘.
I think 27 daily tests is probably far too many to begin with - this might explain the author’s 85% failure rate…. but then again, think of the glow of satisfaction he’ll experience when he passes all 27!

I think this appeals to me in the same way that GTD does - one real lesson from which is to keep taking the small steps….

Found via Phil Wilson’s blog.

Technorati Tags: GTD, personal unit tests

Just came across A Periodic Table of Visualization Methods - not all the ‘elements’ have examples, but I like the idea.

Technorati Tags: visualisation, periodic-table