After an amazing Hackathon, where I had a chance to mine for text and images in web archives and several weeks of working through weekends on various commitments (i.e. co-editor responsibilities, manuscript revisions and the like), I fell sick. The silver lining in not being able to look at a screen for more than five minutes was that I had time to think about what I currently have and where I am going with MINA.
What I have:
(1) Geocoded archaeological investigations between 1953-1960
CSV with additional attribute information on the the nature of investigation, lead archaeologist and team, their institutional affiliation, the culture-historical period of the material culture, and reference to the original publication. I may go ahead and geocode the 1990s set that I have as well.
(2) Git repository of the Digital Atlas of Egyptian Archaeology
Taking Ethan’s suggested workflow, I have downloaded the DAEA repo and I will use it as a template for MINA.
I like DAEA’s clean and simple layout. I have started playing with its ‘look and feel’ and its functionality. Specifically, it draws on point data from a CSV and uses Mapbox (via Leaflet) for map tiles. MINA will have an identical setup.
The DAEA map is centred on Egypt, and uses Arabic for geographic terms. archaeological sites are indicated on the map using blue point markers. Users interact with the map by clicking on the point. A click will result in a pop-up window with a description of the site. This is relatively straightforward when there are a few, well-spaced out markers. MINA will differ in geographic scope, of course. I also know that there a lot more points that will make things messy, so I will likely need to try out a few things with clustering symbology.
I also would like to have additional map interactivity. I want users to be able to filter the CSV based on say, one institution, or a particular archaeologist. The same could be applied for culture-historical period or any other attribute in the CSV.
(3) Git repository for MINA
I have setup the MINA repo here. There is very little there at the moment, other than a landing page. There will be more!
I’ve been experimenting with Mapbox, particularly the default symbology and layer control. The overview will be national i.e. the whole of India (zoom level 5) which looks OK. But in the default zoom level 5, I can’t see any rivers. I realized that the waterways layer in Mapbox is locked for a high zoom level and that means rivers are visible only as you zoom into a smaller area.
I decided the best way to show rivers was to make my own geojson. I checked the metadata documentation on Open Street Map and used overpass-turbo to grab the relevant waterway (=river). Of course, the simple way was to use the metadata to filter and select all rivers, which worked. But the resulting file was too large to process and download on a browser. I decided to get only the nodes for major rivers and this made download way (way, way, way) easier. On Mapbox, the resulting geojsons look decent, but I can see a few gaps. A little tweaking will set it right.
And here is the same shot for rivers in southern India.