Contributor(s): Quentin Devers
It includes the creation of both the geometry and the attributes. For example, if you consider a layer for the streams of Tibet, the task is to: 1) draw the streams, 2) gather the names of the streams in Chinese and in Tibetan from the existing maps you can find.
These datasets can have two destinations: the Place Dictionary, and the Interactive Map interface. To make the distinction, in the first case we will talk of *shapes*, and in the second case of *layers*.
*shapes*: theses are datasets aimed for the Place Dictionary. What does it imply? -> The Place Dictionary (abbreviated PD hereafter) is made of two distinct modules:
What make the link between the two modules are the FID of the places: each place created in the PD has a FID, and then this FID has to be given to the geometry of this place in the postGIS database to have the two associated together. The particularity of this system is that you can have more that one shape in the postGIS database to which you can give a particular FID in order to associate more than one geometry to a particular place. One practical example could be that for the USA, you can have one polygon showing the USA at the very beginning of its creation after independence, then one polygon showing the USA each time a new state is added. In this case, we would have a set of polygons that all have the same FID, and they would all be associated with the single entry in the PD called "USA".
On a practical level, you're going to create the shapes in a GIS software such as ArcView. For the attributes the most convenient is to create a spreadsheet. And to have a FID colon linking the two (on a first step you probably won't have a FID, so just take any kind of made up ID system that you will transform into FIDs later on). To know how to get FIDs, ask Andres Montano.
One wiki about how to import shapes to the postGIS database: [THL Place Dictionary Shapes Importation Manual|https://collab.itc.virginia.edu/access/wiki/site/c06fa8cf-c49c-4ebc-007f-482de5382105/thl%20place%20dictionary%20shapes%20importation%20manual.html]
*layers*: these are datasets that are not aimed for the Place Dictionary. For example, if you have a shapefile showing the literacy in the different Tibetan counties: such a shapefile is not made to go in the Place Dictionary. These are not just shapes (in the PD the shapefiles you create are just shapes without attributes that you upload to the postGIS database), but these are *layers* with specific attributes. Such layers are aimed to be visualized in the [Interactive Map|http://thlib.org/places/maps/interactive] (under the "Map Layers" section). Two things about these layers: 1) you need to make the names of the different fields of the shapefile clear and easy to understand, 2) you have to create a symbology that makes it easy to understand what one is looking at.
One wiki about how to add layers to GeoServer: [Importing SHP Files to PostGIS/Creating GeoServer Feature Types|https://collab.itc.virginia.edu/access/wiki/site/c06fa8cf-c49c-4ebc-007f-482de5382105/importing%20shp%20files%20to%20postgis-creating%20geoserver%20feature%20types.html]
In a general way it's good if you have a look at the different wikis contained in [THL Toolbox > Places & Geography|https://collab.itc.virginia.edu/access/wiki/site/c06fa8cf-c49c-4ebc-007f-482de5382105/places%20%26%20geography.html]
To do the job, you'll need to have logins and write/read authorizations to these different systems: