The data structure presented here represents a hybrid approach based on best practices identified in OCHA offices around the world. It includes standards both for where data is stored, in what format, and how it is named. Additionally, it provides structures for archiving outdated datasets and products.

Folder Structure

The primary goals of the folder structure are to

• Provide a predictable location for datasets, production files, and output products at country, regional, and global levels
• Make data and products easy to find
• Avoid breaking data links (such as within MXD files) when data is updated or copied from one computer to another
• Provide distinction between "finalized" datasets that are ready to be used in cartographic products and datasets that are still in production
• Provide a space for archiving outdated datasets and cartographic products
• Provide a space for working on datasets or products separate from the final location of the output files
• Handle data in multiple formats (Geodatabase, Shapefile, KML, GML, etc.) as well as production files (MXD, DOCX, AI, etc.) and output files (PDF, JPG, etc.)

A note about existing folder structures: Most OCHA offices have evolved their own folder structure over time. This manual does not intend to impose this structure on any office. As offices review their folder structures or as new offices are opened, this structure is recommended as it satisfies the goals listed above. In particular, the highest level folders (the 1st and 2nd levels below the root level) are critical to enabling data synchronization between country, regional, and global offices.

Two tools have been developed to help explain the system:

• A zipped empty folder structure, using Afghanistan as an example  (NEED to UPDATE THIS)

Archiving

Notice there is no date on dataset names.  This allows us to update data files without breaking links to the MXD files. To archive a file using this approach:

1. Give the outdated file a date stamp at the end of the file name.  Use the date it was removed at the end of the file name in yyyymmdd format.
2. Place the outdated file in the co-located ‘archive’ folder.  
3. The new dataset should have the same name and location as the previous (now archived) file.

Any project files referring to this data layer will then automatically use the new layer.  

For example: 
AFG/Data/Boundaries/afg_polbnda_adm1_1m_salb.shp  
is archived on March 13, 2011 as 

AFG/Data/Boundaries/archive/afg_polbnda_adm1_1m_salb_20110313.shp  

The updated file will be named 
AFG/Data/Boundaries/afg_polbnda_adm1_1m_salb.shp  

Working Directories

Any folder can have a working directory.  It's just a place to make a mess when you are working on something.  When finished, move the file (whether it's data or a product) to its appropriate place and clean out the working folder.  If there are working files you want to keep, put them into a well-named folder and store them in the appropriate Archive folder.