Showing posts with label Geospatial data. Show all posts
Showing posts with label Geospatial data. Show all posts

Thursday, 21 January 2010

Digimap is 10

I had a very enjoyable day yesterday helping EDINA celebrate 10 years of the Digimap service. What began as an eLib project and experiment with 6 Universities in 1996 has grown to a mature service with over 100,000 users, 45,000 of them active, in pretty much every UK University, and soon in UK schools as well.

In 1996 I was Programme Director of the eLib Programme, and my earliest email about Digimap was from the JISC money man, Dave Cook, on 30 January 1996 to Peter Burnhill of the Edinburgh Data Library (as it then was). Dave told Peter we were interested in his idea (for an Images project!) but had a few concerns (that the Ordnance Survey might not agree to let us use their mapping data; it’s hard to remember now how difficult some of those 1990s persuasions were!). Three days later, Dave was offering real money, although it had to be spent by 20 March that year. Done!

By late 1997 the Digimap project (*) had a trial service; I remember experimenting with it and having some problems (this was with Netscape 3 on a PowerMac Duo or something like that; woefully under-powered in retrospect). By the end of 1999, they were moving to a new GIS system, and we were beginning to discuss turning Digimap into a service, and that went live in January 2000. They had to get 37 subscribing Universities by a particular deadline, and I think managed 39 by somewhat earlier.

Since then the service has grown in scope, quality, usage and value. In my personal opinion (full disclosure, I’m not neutral here, having been associated with it through advisory groups of various kinds throughout its life), Digimap is the best service funded by JISC. Best in quality, best in professionalism, best in innovation, best in support. A lot of people deserve credit for that, and EDINA should all be extremely proud of what they have created. By the way, the OS have managed some major shifts in attitude over the years, from suspicious tolerance through to strong support, and the success is partly down to them, and to the efforts of the negotiators in what is now JISC Collections.

As well as various forms of OS mapping for GB (whose trademark names always escape me... and it is GB rather than UK, for weird historical reasons), Digimap now offers 4 “epochs” of historic maps from Landmark, plus Geology maps from BGS and Marine maps from SeaZone. Due to licence restrictions it is only available to registered staff and students at subscribing UK institutions, but I hope that those of you unlucky enough not to fall in that category can soon read more about it on the pages to be put up related to the celebration.

Digimap has been a bit clunky at times compared with the innovations introduced by some others, but with the new underlying GIS, the interfaces are being upgraded; they now have “slippy maps” (called Digimap Roam) on the base service, and it looks really smart and much more functional. It's tough for a small group to keep up with the likes of Google, Yahoo and MS! Soon this slippy map interface will be extended to the Historic service (“Ancient Roam”?), Geology (“Rock’n Roam”?) and Marine (a rather dull “H2Roam”!)… I think those might be internal names, but if you can complete the set with an even punnier marine name, who knows they might keep them!

The day was good fun, and we heard quite a bit about what Digimap is and how it is being used (far more widely than geography departments). The most exciting was a student project using Digimap and a GPS for a light aircraft CFIT-avoidance system (CFIT is Controlled Flight Into Terrain, referred to as “having a bad day”!). We heard from the data suppliers, with a bit more about what’s coming. It was interesting to hear the OS man talking about moves towards Linked Data; I wasn’t sure how that would square with the closed access, but I think I muddled my question (confused Linked Data with OGC web services, I suspect). The service providers didn’t appear to be talking to each other about Linked Data, which might be a good start.

A highlight was the closing keynote from Vanessa Lawrence, CEO of OS, clearly extremely supportive of OS. Choosing her words very carefully (she is not allowed to influence anyone) she outlined the government’s open data initiative and the consultation on its implications for the OS; this consultation closes late March 2010, but she urged us to make any responses, whether collectively or as private citizens well before then. The consultation isn’t simply “should we open up access to OS data?”, it’s much more “how can we open up access to OS data and still sustain the quality of the data into the future”.

The celebration ended with a reception and dinner, with an amusing after-dinner talk by Michael Parker, author of Map Addict. All in all, a very enjoyable and worthwhile day to celebrate a significant anniversary.

PS the twitter tag is #digimap10; I’m not going to tag the post with it, as I’ve got far too many one-time tags that are a pain to manage…

PPS (*) Unfortunately the original Digimap project pages seem to have vanished, and the earliest Wayback Machine gathers appear to be faulty; the first successful gather I can find is

http://web.archive.org/web/20011021051021/edina.ac.uk/digimap/

... which seems to refer to the service, not the project.

Tuesday, 9 December 2008

IDCC4 afternoon research paper session

I really want to write about the afternoon research papers at the International Digital Curation Conference, but I start from a great disadvantage. First, I wanted to hear papers in BOTH the parallel sessions, so I ended up dashing between the two (and of course, getting lost)! Then my battery went flat, so no notes, and I left the laptop in the first lecture room while going off to the second… and of course, near the end of that paper, someone came in saying a couple of laptops had been stolen. Panic, mine was OK, thank goodness (when did I last do a proper backup? Ahem!).

I’ve reported on the Australian ARCHER and ANDS topics before; the papers were a little different, but the same theme. The paper I really wanted to hear was Jim Frew on Relay-Supporting Archives. Frew’s co-author was Greg Janee; both visited the DCC, Frew for a year and Janee for a few days, talking about the National Geospatial Data Archive. I thought this was a great paper, but I kept thinking “I’m sure I said that”. Well, maybe I did, but maybe I got the ideas from one or other of them! They have a refreshingly realistic approach, recognising that archive survivability is a potential problem. So if content is to survive, it needs to be able to move from archive to successor archive, hence the relay. An archive may then have to hand on content that it received from another archive, ie intrinsically unfamiliar content. How’s this for a challenging quote: “Resurrection is more likely than immortality”! He looked at too-large standards (eg METS), and too-small standards (eg OAI-ORE), and ended up with the just-right standard: the NGDA data model (hmmm, I was with him right up till then!).

The conclusion was interesting, though:
"Time for a New AIHT
  • Original was technology neutral
  • New one should test a specific set of candidate interoperability technologies"
The conference closed with an invigorating summing up from Malcolm Atkinson, e-Science Envoy. Slides not available yet; I’ll update when they are. He sent us off with a rousing message: what you do is important, get to it!

See you next year in London?