World Geo-History Data Archive ============================== Overview -------- The purpose of the data archive is to provided a wide-reaching interactive map-like view of Earth throughout it's history, in both human and cosmological scales. Rather than be an encyclopedia in it's own right, the data is meant to form a kind of giant interactive index structured by time and space, and how things in that time and space interact, rather than an alphabetical list of everything in existence. Layers of data should range from the movement of continents over millions of years, to the few millenia of human civilisations spreading and retracting across the globe, to the ability to pick out where an invention occurred or a historical figure was born. The project was born out of my realisation that my view of the overall flow of the world was wrong, even after my attempts to correct some of the biases in my historical knowledge from my Western European (specifically English) education. The current example of this that comes to mind was my complete failure to realise that Christianity was present in East Asia during the fall of the (Western) Roman Empire, or my later lack of understanding why in Europa Universalis 4 (a wonderful game to look at semi-historical world maps over 400 years) Indonesia was defaulted to Islam (the Muslim period of Indian Ocean trade as the main factor is my current understanding). The initial design is intended to be limited to the planet Earth since becoming spheroidal up to the present day, incorporating method for all storing all notable historical and geological data. Key Data Storage Concepts - Entities ------------------------------------ The key thing to realise in this endevour is that attempting to break down history into periods, era, or ages does not really work when attempting to view the entire planet on a range of timescales -- each part moves at its own pace. The canonical example of this for British taught pupils will be the European Dark Ages which were not especially dark, have no clear beginning or end, and only apply to (parts of) Europe. The slightly larger example is the extinction event of the dinosaurs, which was primarily precipitated by a meteor impact, but still took longer to occur than the current complete history of homo sapiens. As each person, city, domain, and civilisation rises and falls by its own history (and interactions with neighbours), it make sense to track each unit of historical interest as a separate entity across the time of its existence. Each of these examples generally has complex relations to entities of other types -- cities are often parts of domains such as kingdoms, which in turn may represent the entirety or part of a civilisation. Furthermore, there will be the concepts of (diplomatic) relations, leadership and other inter-entity relations, varying over time. Key Data Storage Concepts - Data Points --------------------------------------- To all these end, every entity that is part of the dataset for the map will have a series of data points or events associated with it, each of which is tracked independently of events on other entities. Each data point will boil down to four primary properties: - A reference to the entity to which it references; - The type of the data - The content of the data point itself - The translation relationship A simple example might be the population of the city of London. The first three data fields for a few data points might be: { "London", "1931", "4,900,000" } { "London", "1939", "4,440,000" } { "London", "1951", "3,680,000" } (These figures are for 'Inner London', from [1]). By taking the relevant data points of all entities, it should be possible to reconstruct an accurate representation of the world at the time. However, by having discrete data points, we have to find a way to interpolate for an arbitrary date. This is the purpose of the fourth main property of the data point, the transition relationship. Basic examples of these relationships could be: - linear: change between the two values is the same over time - exponent: change between the two values follows an exponential curve - snap: change between previous and new value is instantaneous at the event. The different types are generally more common is entirely different contexts: 'snap' changes work well for borders and land claims, when declarations change the state of the map. Data Points - Location ---------------------- One of the key common pieces of interest both in geology, geography, and history is where things happen. When considering geology and history at the same time, this leads to the major issue of trying to put pins in a map where the borders are ever so slowly (continentally) drifting around. There is also the minor matter that the Earth is some kind of oblate spheroid [citation needed?]. Until such a time that extraterrestrial travel is is common place, there is little need to deal with the interactions of Earth and other Celestial bodies. However, the three dimensional nature is important to a lot of modern life, such as why planes don't fly in straight lines over the maps we produce. In order to be able to represent the overall map in both 2D and 3D spaces easily, the intention is to use a high precision longitude and latitude values, with an optional altitude above mean sea level. This can then be combined with a model of Earth of any detail, and some basic trigonometry to present a 3D version of the model. Entity Types - Regions ---------------------- Meh. Bored of typing now. Footnotes --------- [1] http://www.londonspovertyprofile.org.uk/indicators/topics/londons-geography-population/londons-population-over-time/