Defragmenting Digitized Manuscript Sources

A unified portal to medieval manuscripts

Let’s begin from the core elements: The DMMapp is based on two freely available products: Google Fusion Tables and Google Maps API.

Google fusion tables is defined by Google in the following way:

an experimental data visualization web application to gather, visualize, and share data tables

while the Google Maps API allows anyone to build and design a map using JavaScript.

Let’s give a look at how these two interact and how we insert a new library into our database.

Imagine that a user has just compiled the form available on the DMMmaps website, in which he or she has entered the name of the institution that should be added to the map, the location; the number of digitized manuscripts available on the institution’s website, and the link to these manuscripts.

The first thing that we do is visit the given link, and check if the information is correct. If it is, we copy-paste the given information into our spreadsheet, into the proper columns: City, Nation, Library Name, Website Link. We then find the longitude and latitude of where the hosting library is using Google Earth, and also copy-paste these coordinates into the spreadsheet.

Finally we give a look at the copyright with which the images are protected, indicating whether Creative Commons is used or not, and, in case, under which specific CC license.

This same exact data is then sent to Google Fusion Tables. Fusion Tables reads the coordinates that have been inserted for each library and creates a very basic “pin” on a map, that, when clicked upon, shows all the information that we have collected for that particular library.

It must be mentioned that this “Fusion Table” is publicly accessible and usable; anyone can go on the fusion table’s website and find the same exact data found on the final DMMapp.

So, in theory, the project could be considered complete at this stage: it is accessible, it is usable.

But we decided to go beyond. We decided to create a web app that was not only useful, but also visually appealing and easy to use.

This is where the Google Maps API comes into play. While only part of the design features are editable in Google Fusion Tables, the Google Maps API allows us to edit every aspect of the map.

Now, I have mentioned the word “design” a a couple of times already. This is an aspect that I believe to be extremely important within any project, including a Digital Humanities one. Often, extremely interesting projects lack a little bit of visual appeal, or user interface design that would make using such project so much easier and faster for everyone.

In our case, we redesigned the way information presents itself to the final user: The name of an institution, the nation, and the city where it can be found are now on the same line. The simple link has been replaced with a clearer text, and the number of available manuscripts is detached from the rest of the information. The box that appears when a pin is clicked upon, and where all the information is, is also more spacious and pleasant to the eye.

Moreover, the Sexy Codicology blog uses an immediately recognizable color scheme: blue, white and grey. We wanted these colors to be present also in the map, to seamlessly connect the blog to the map.

The final step is the webapp interface. We developed this with both desktop and mobile users in mind: We wanted it to be usable and look good on any screen.

We therefore based it on Bootstrap, a framework for developing responsive projects that look great both on desktop PCs and on Mobiles; and Searchable Map Template by Derek Eder as a starting design point.

We edited some code, and this is what we created:

In this slide you can see the DMMapp in action: We have zoomed to Madrid to give a look at what is available there. In this case we can see that the National Library of Spain has 339 digitized manuscripts waiting for us. All we have to do now to access them is to click.

Notice also the social media sharing buttons on top, and the “follow us” button for Twitter and Facebook. All clearly visible and easy to reach.

Now, the name of our project is DMMmaps, and not DMMmap. Up until now we have shown only one map, but it is time to show also another side project we have been developing: The Digital Libraries Heatmap, which you can see in this slide, or access from our website or app.

In this map, every library added is a heat-point. The more heat-points are close to each other, the redder the color becomes.

This map is develop to show at a glimpse of an eye how nations are performing when it comes to the digitization of medieval manuscripts. In example, while France, Germany, and Switzerland are doing wonderfully, southern Italy and Spain have only very few digitized libraries.

I have mentioned the spreadsheet a few moments ago. This is also publicly available and linked from the app and the website. Anyone can consult it, download it, and use it to create their own project.

We are also very active in blogging about the libraries we visit and that are available on the DMMmaps. In the posts we analyze some features that we believe to be important in any library that makes digitized manuscripts available on internet: navigability, image quality, and copyright protection. We also use these posts to promote the contents of these libraries: we highlight the treasures that can be found on the website, explain why they are important, and link directly to them. In these few months we have already sent several hundreds of visitors to various digital libraries thanks to the blog posts alone.

At this point, it is important to mention that the DMMmaps is a privately funded and managed project. This means that the project has extremely limited financial resources that can be invested into it.

Every year we will invest 18 euros in fees to maintain the hosting and server space available for our app. This is a cheap price, but with a cheap price comes also the fact that the hosting company we rely on is not the most stable. In these past 5 months we had several occasion in which our server went completely offline, rendering the app and the blog unaccessible. The previously mentioned spreadsheet, on the other hand, is always available since it is hosted on Google servers rather than ours.

I can’t see a way to monetize the project in order to improve its infrastructure. For now, the little income coming from donations (60 cents so far!), ads (two cents a day on average), and the Sexy Codicology merchandise (around 16 euros in two years), helps mitigate our expenses.

Beside this structural weakness, there are also other issues that affect the project’s base itself.

As I have mentioned, we try to add how many manuscripts are available per each digital library, but this is not always possible. Often the number of digitized objects is simply not displayed on the website of the digital library, and there would be no other way than counting how many manuscripts there are in order to add that number to the DMMmaps. I could be done, but I believe time is better spent otherwise. Moreover, these numbers are subject to constant change: it often happens that a repository has five digitized manuscripts on a month, and a month later these have become 20. It is clear that keeping track is difficult, but we do our best.

We are also not able to make a distinction between fully digitized manuscripts and a partially digitized ones. Many times institutions don’t mention whether the full book has been made digitally available or not, and, to make matters worse: sometimes fully digitized books go together with partial ones, or mixed with other media that cannot be filtered out by server calls.

Another big issue are broken links. Websites change all the time, and often do web addresses too and the same goes for websites that host digitized manuscripts.

To make an example: recently the digital manuscripts of the Vatican Library have moved to a new address. The old web address we linked to was still accessible, but it would not redirect to the new location and showing a nice empty list, leaving the visitor confused. We went to the website, found the new link and updated the DMMmaps in five minutes. No big issue, but proper web-mastering on behalf of housing institutions would be useful to all parties.

Although this issues, we believe that we managed to make a good product; one that ended up being just like we thought it should have been when we first started: easy and fast to update, easy to use, easy to find on the internet, cheap to make, and, most importantly, useful to anyone who is interested in finding digitized medieval manuscripts around the world.

We will continue ironing out the small issues that can be found here and there on the app, continue adding and reviewing libraries with the help of everyone, and, maybe one day, develop an real app for smartphones and tablets.

I hope I have exposed how the DMMmaps work clearly enough. We will be giving a demonstration later on, in case you are curious to see how everything works in practice.

Thank you very much for your attention.

Marjolein de Vos, Co-Manager and Web-Editor at Sexy Codicology, opened the presentation speaking freely while looking at the following slides.