On Wednesday I attended a JISC meeting on the subject of Persistent Identifiers (#jiscpid) which is a long term topic of discussion in JISC-world. Now I came to this as pretty much a newbie. I have a long history of obsessing over the design of URL schemes and remember reading the original(?) TBL paper on Cool URIs years ago but I have never really been concerned with the wider picture as regards things like digital preservation until quite recently – previously (and still really) I was more interested in creating human hackable URLs.
[On a bit of a tangent my favourite URL scheme I have seen in a long time was discussed in this article over on Carsonified – I am also a big fan of the http://traintimes.org.uk URLs – great stuff and an inspiration I think..]
In preparation for the meeting I spent alot of time reading up on same of the schemes and technologies involved in the ‘persistent identifier’ space and I’m glad I did or I would have been totally lost. The issue seems to have risen to the top of peoples agendas recently due to the rise of Linked Data and the role of things like ‘cool uris‘ and identifiers within that movement. I am reasonably familiar with this activity so I focused on the alternate universe that exists just out of step with our own. A world of DOIs, Handles and ARK. A world where publishers and researchers have created solutions outside of the normal web architecture to deal with very real issues but in a manner that will (it seems to me) never break out of these very tight use cases around (scholarly) publishing because it is just too different from what most people find familiar. That said these solutions cannot be dismissed as they have created solutions for their users and built businesses and communities around these solutions.
One of the problems (though as it turns out far from the only one) ‘persistent identifiers’ seek to solve is the very real issue of ‘linkrot‘ which has long been a thorn in the side of any serious user of the web – and a particular pain in the rear for me for a large part of my career. The thing here though, as was mentioned multiple times, is that for the most part this isn’t really a technical issue but rather a political and organisational one. Websites are re structured, departments renamed, platforms updates. All of these activities often cause problems to persistence. However DNS and server technologies already offer the tools to mitigate these happenings if appropriate resource is given to the task. This is about education of management as much as anything. As one of the attendees at the meeting said the website is not seen in terms of how it performs as an information resource for the wider web community but how it responds as a marketing tool and thus persistence is not as important as the next big thing.
Something I found interesting was part of the work of the RIDIR project which had created a kind of ‘crowd-sourcing’ solution to fixing ‘linkrot’ for Digital Repositories by allowing people to associate the found resource with the no longer working ID or URL (this is based on the – I think – very real use case that when a link is found to be broken users often perform an alternative search to attempt to find the original resource – once they do this they map it using a ‘widget’ or browser plugin to the broken link creating a redirect essentially. The system allowed for multiple endpoints to be mapped if there was a disagreement and for it to be left to the users discretion as to which option to follow based on the reputation of the original mapper. No idea how something like this could be scaled for the web in general but I think it would be interesting if it could avoid the inevitable spambots!).
All-in-all I came out of the meeting without the bolt of lightning, moment of clarity I had hoped for on the subject but with a real sense that this was an area of importance and that maybe it was also an area where, to paraphrase someones tweet, there was a sense of compromise and pragmatism that could make something happen. I also realised, not for the first time, that I am forever destined to see the world through the eyes of a Web Manager and I will always boil issues down to things I can understand in that context.