23 years in the making
I trained as a Graphic Designer. But back in 1995 I graduated the first part of my Masters, a Post Graduate Diploma in Design for Cultural Databases and I’ve worked in design for digital media and technology ever since. However, I’ve never worked on a ‘design for a cultural database’, well that is until 18 months ago.
As my other blog paperposts.me reveals I still have a huge passion for all things graphic design and around two years ago I had my first visit to the Letterform Archive in San Francisco. It’s an incredible collection of graphic design, typography and calligraphy from books and periodicals to ephemera, objects, posters and other oddities (more about my discoveries there). From the moment I visited I fell in love with the place and people. The knowledge and range of items is amazing, especially as they have a lot of sketch, prep and process work accompanying the designs. And it’s massive too — today they have over 50,000 items as well as 60,000 Linotype drawings! From that first visit I knew I wanted to get involved.
At the time I’d been playing with an approach for viewing, analysing and filtering images. I’d used it on a few public datasets, like the ones issued by the New York Public Library and Tate Modern and I thought, ‘wow, it would be amazing to explore the Letterform Archive like this’. So I went looking for some data. Alas there was none as the Archive had only been up and running for a couple of years and the digitisation work was just beginning. However, they did have a fairly rich instagram account, so I set to work on that.
I’d done a few volunteering sessions with them and got to know the team quite well. After a couple of months I decided to share what I’d been up to. Basically I’d downloaded most of their IG images (just over 1000) and gleaned as much data as I could from the comments. Pretty messy stuff but it had the key thing — the images. I’d pulled keywords, dates, hashtags, people etc but no truly structured data. I’d also taken a stab at analysing the dominant colours too. It was a lot of fun.
In early 2017 I met with Amelia Grounds (the Archive’s librarian) and a few others to share what I’d been playing with. There was a lot of excitement but also some cold hard reality about where they were in the both the cataloging and digitisation process. The core digital system of record (including the software), had yet to be defined and any imagery shot had just been for social or editorial reasons — essentially there was no structured digital archive. There was a road map, but it was at least 5 years out before any type of explorable online catalogue, let alone consumable data for 3rd parties to play with.
But where there’s a will there’s a way.
As the digitization was yet to start we had a huge opportunity to build in the data needs from the start. So many digital experiences for archives are hampered by legacy cataloging systems and horribly messy data, we had a great opportunity to tune the archive for explorable data, rather than simply a system of record.
We were hooked, seeing the opportunity to create something truly wonderful. We spent the rest of 2017 looking at data structures, MARC records, approaches to handling images and everything we could think of regarding the data. I’m not a data scientist, or a particularly strong developer. I’m (at my core) a designer — so I spent a lot of the time learning and experimenting. (And I’d just like to add that MARC records are horrible things and I feel for all those data people that have to work with them.)
What we arrived at was a very small proof of concept that had images, structured data and could be explored. Essentially a way to ‘browse’ the Archive (even if that was a very thin slice of it). We were pretty chuffed and felt it was time to pitch it to the rest of the team, in particular to the Archive’s founder — Rob Saunders.
At this point we had created structured data for 127 items that had a corresponding image on IG. All this data was crafted by hand by Kate Long, with lots of cleaning and wrangling. I then took it exploded it and stitched back together to create an associative data model that would allow people to filter the items and get a sense of where those sets overlapped. It was all put together using the free version of Qlik Sense Desktop and a custom extension I’d made. The surface experience was simple; search and filter the images. But underneath we had started to craft a very rich data platform. Joining a bunch of separate in house cataloging spreadsheets with IG records and MARC record data from the Koha open-source integrated library system (the Archive’s newly selected system of record).
This pretty much blew the team away. Rob immediately saw the potential and the possibilities for an online ‘discovery’ tool that would enable those that couldn’t visit the Archive to get a taste of what it had to offer. It was also one of those “Jaws zoom” moments when Amelia and I realised how much we needed to do to actually deliver this. It needed a budget, data, images, a proper platform and real technical development, a designed experienced tailored for those who’d actually be using it and of course — a plan to do all this. Basically, ‘this shit just got real’.
Since then almost everyone at the Archive has been toiling away in the background helping to build this. Today thanks to Amelia’s stewardship and the herculean efforts of Kate Long and her volunteer catalogers, as well as Rachel Daniels’ camera work we have over 3600 items digitally catalogued, with over 6000 images. With the help of the rest of the team (not least Stephen Coles) and Jon Sueda, chris hamamoto, and Omar Mohammad who supplied some front end design love, this month we were able to release a private beta for Archive members. It has just over 1000 items with approximately 2900 associated images and is very much the beginning. Helping us learn what we need to make this a fully public experience and how we can give members a virtual experience that reflects and supports their physical visits to the Archive.
So I’m very proud to say that after 23 years I’ve finally designed and delivered a cultural database.
Here’s a preview.
There’s lot’s more to come, so JOIN THE ARCHIVE
UPDATE 2020/04/08
As of yesterday, with special thanks to Nick Webster, his Qlik Core skills , his rebuild of my shonky front end and the Qlik integration, it’s live to the public:
http://oa.letterformarchive.org/
Enabling anyone, anywhere to view 1473 distinct items from the collection:
Rebuilding the experience so that it scales for public access:
More about the launch:
