dev8D

Cool code: EPrints export plug in allows preview magic

Developer Happiness Days has seen a number of immediate and exciting ouputs from the industrious coders here. While Sam Easterby-Smith’s Happier Pipe is arguably the most fun, Chris Gutteridge’s speedy work on JSON exports for the EPrints software should have a real impact on the ease of use of the popular EPrints service. Chris explains:

“I have built an export plugin for the EPrints software - http://eprints.org/software/ - which allows records and searches to be dumped as JSON. This will allow other Dev8D delegates to built new tools on top of this data.”

JSON makes it much easier to use your data in mashups on third party sites. Examples already demoed at dev8D include a third party website which allows users to browse the data in a new, faceted way, and a  little code to add to the top of a webpage, which makes previews magically appear on each eprint link. Try hovering the mouse over links on
http://www.ecs.soton.ac.uk/people/lac/publications

To access EPrint records as JSON, either perform a search on  http://eprints.ecs.soton.ac.uk/  and select the export as JSON option; or as a “REST” URL for record 15818:
http://eprints.ecs.soton.ac.uk/cgi/export/15818/JSON/ecs-eprint-15818.js
or,  Chris has dumped all the data at   http://eprints.ecs.soton.ac.uk/dump.json (warning: 35 Meg)
There is also a  YAML version.

I have an idea! Part 5

Julian Cheal
UKOLN at the University of Bath

My idea is called JISC Conferenceator. I came up with it yesterday as this week everyone has been aggregating all these different feeds together and having a conference backchannel using Twitter and creating friend graphs so, inspired by that, the JISC Conferenceator is a toolkit which would bring all this together on one site. You would create your conference on this site and bring together all the facilities like built-in support for surveys, an email address and do automatic RSVping and connect to open social networking sites and Flickr so users can connect to that, and all the data would be brought back into the JC site. It will also make it easy to capture delegate feedback – much better than having a long form. All the data will be in one place so after the event you can zip it up into one file to keep.

I have an idea! Part 4

Andrew Wasilczuk, Darq (an open source networking consultancy)
Ian Morrison, Darq

Our idea is based on the premise that adverts are bad on websites as they are a waste of bandwidth and screen real estate – they take up room which could be better used. We want to repurpose that space and filter out the adverts and replace them with relevant content. We would implement it by replacing things like Google ad words with relevant data that’s helpful to the university member. So if a student goes to Google and does a search for, say, context switching, Google returns its normal results but, instead of ads on the right hand side of the screen, there would be a box containing local resources. Local in this sense is local to the institution that the user is acessing it from eg links to local library and books relevant to the search, student forums, course notes from the intranet and so on.

Google is very relevant but it is relevant on a global scale – we want to make it relevant on a local scale. An application server, a widget, will generate these local frames, inject them into the page and that widget talks to a sever which knows the answers – a local semantic web. At a simple level we take out the adverts and replace them with something else. What that something is depends on what there is a need for.

It could be done with a Firefox plug-in but then it becomes something else a user has to configure so, to make it easier and remove deployment issues, we would do it at the network level with a transparent proxy (a machine that sits between you and the internet and looks at the traffic and removes the adverts and injects our content). It would happen automatically if the user is connected to the university network.

I have an idea! Part 3

Marcus Ramsden, University of Southampton

I’m working on a small script plug-in for e-prints. The ultimate goal is for it to be a firefox extension so that whenever an e-print link is displayed on screen, it will display information about that e-print before you click on it. It will be basic meta info – title, authors, date it was added, basic statistical views. In future it could be extended and show things like how many people favourite that e-print etc. It would have been much trickier to do this without the JSON addition created here at dev8D by Chris.

Graham Klyne

I did have a possible idea for the Developer Decathlon which would involve some paper prototyping of an interface for capturing research data from small research groups. However, in the process of following that idea I had a discussion with another participant and he showed me a public service which does a lot of what I had in mind so it seems that we could use that service as a live prototype to discuss with researchers. So I regard that as a very positive outcome.

Stephen Vickers, University of Edinburgh
Michael Aherne, Strathclyde University

It’s a tool for using within VLEs that allows people to create things by plotting points on Google Maps and associating things with the points - it links spatial data with content. The original idea came out of a project on walking tours but the data wasn’t so easy to access and plot then. Now both staff and students can do it so we plan to use it for assignments, too. We’re looking at a history department and the impact of the urban landscape over time, putting things in the context of what happened over history. We can overlay historic maps onto the map and see how it has changed. It will help people to understand why things have changed. It doesn’t have to be a map, it could also be an image like a forensic site or archeological dig or a circuit board.

I have an idea! Part two

Mark Dewey, UKOLN
Chris Yates, RSP
Stefan Secheud, Centre for Digital Library research at Strathclyde University

It’s a semantic search. It works on a weighting system. When people sign into repositories we can track what they are searching for and what groups they are part of. It would gather different bits of information about the user and build a search index and target anything they would later search for against this weighting which had been previously determined. It would give them more targeted results. Users frequently complain that repository search is not particularly relevant. It comes up time and time again that search terms don’t bring much relevance so this should make academic searches more useful to people.

Juliet Culver, OU
Matt Zumwalt, MediaShelf
Edwin Shin, Fedora Commons and MediaShelf
Chris Wilper, Fedora Commons

We spent yesterday chatting to Dick, a biology lecturer here at Birkbeck and he wanted his students to write better lab reports so we had a chat about how that could be improved. One of the things is that they want them to produce drafts and then get feedback and iterate their lab reports to produce better ones. We want to come up with a system that provides ways for the students and the lecturers to give feedback on the whole lab report and also on specific parts of the report. It would also direct people to help eg tips and checklists and direct them to other reports to look at for various sections. If you can give students advice at a particular point in the process then you can help them to avoid those problems in future. The metric for success is that students will write better academic papers so we hope it will not just be the instructor doing to commenting but also open it up to their peers. It’s about doing more than one draft so that people are learning from what they do. This could also be adapted to other disciplines.

Dev8D produces rapid results

Day three of Developer Happiness Days is only just beginning but two ideas have already been made real by the keen coders here.

Splash URL came out of a plea by Tony Hirst in his Mash-ups talk when he bemoaned the lack of an easy way to shorten a long url and have it appear in large type in the centre of the presentation so that people can easily copy it down. No sooner said than done - Chris Gutteridge jumped on the case and Splash URL was born.

According to Tony Hirst, Developer Happiness Days is working its magic:

“I’m doubly happy because we’ve got the SplashURL working and I’m really happy that apparently as a result of things that have happened over the last few days E-prints has got a JSON output. This means that the output can be easily pulled into a webpage leading to all kinds of mash-up joy. JISC’s willingness to engage makes me really happy!”

Meanwhile, Sam Easterby-Smith is measuring the happiness of the developers at dev8D in real time using Twitter. Whenever anything is Tweeted using the dev8D tag, if a fraction is included to indicate happiness (such as 9/10), it gets added to the dev8D Happier Pipe and the total sum of happiness at dev8d can be seen at a glance - and in lurid colours, too. Right now, dev8D is looking pretty darn happy.

“There’s nothing like a bit of dirty code cooked up over a nice curry,” said Sam. “It’s been very good to just get to hang out with other developers and exchange ideas, a lot of which have been quite outside my normal comfort zone and what I do. It’s been really valuable.”

I have an idea!

Dev8D’s Developer Decathlon is all about developers talking to each other, talking to users and coming up with great software ideas that they can rapidly prototype. We’ve been been roaming the classrooms and bars here at Happiness Days and finding out what kinds of ideas are floating in the air.

Ross McFarlane, University of Liverpool

I was thinking that in a bricks and mortar library, when you’re looking at content in a particular section you have the opportunity to chat to people who are browsing similar materials as related materials are located together. So my idea is to try to find a way to connect related materials by what people look at online and take out of the library and use that to connect users so that as someone is looking at an item online they can speak to people who have got similar interests as they are looking at similar materials. It’s about creating short term social networks based on the content they are looking at.

Peter Sefton, University of Southern Queensland

My idea comes from a lecturer who wants to be able to assemble powerpoints from resources he has found around the place – rich powerpoints including video and so forth. My idea is to have an add-on for your browser where if you’re looking at a resource you want to include on your course you can click a button and it will remember what you’re looking at and it will store those as tags in Delicious. Then, if you want to make a presentation for your course it will look at the things you’ve bookmarked and build them into a content package that you can extend, annotate, explain, reorder and publish to the virtual learning environment or the web.