Out of the 400+ existing tags, a so-called ‘schema’ can be created that contains exactly those tags that are applicable to a certain type of document (such as a letter that is prepared for a scholarly edition). New tags can be defined when the existing tagset is insufficient. The schema describes the required and permitted tags in a class of XML documents. It can be used to check the correctness of these documents. A dedicated schema was created for the Van Gogh edition. A number of non-standard tags were used, some of which were ‘borrowed’ from the DALF (Digital Archive of Letters in Flanders) project.5050. http://www.kantl.be/ctb/project/dalf/.

One XML document was created for each of Van Gogh’s letters and each related document. It holds letter-level metadata (title, number, date, correspondents, etc.), the full transcription, the translation, the notes, the textual notes, and the information that connects transcribed pages with images of those pages (facsimile elements). The XML files were created in an automatic conversion from word-processor documents. The conversion result was checked and extensively corrected. The XML files were manually indexed to facilitate searching and cross-referencing.

Other data sources for the edition include a number of databases. The editors created databases containing physical descriptions of the letters, data about the illustrations, and information about the people mentioned in the correspondence. Those databases were used for various purposes, such as to support searching and to provide the illustrations with captions.

The facsimile images of the letters were created by the Metamorfoze programme, the Netherlands’ national programme for the preservation of paper heritage.

The letter pages are set up as a row of 1 - 4 columns, each with tabs to choose which section of the letter data to display. Javascript is used to fetch the selected data. Column settings are stored in browser cookies so visitors can use their own configuration the next time they visit the site using the same browser.

We used the ImageMagick suite of tools for cutting up the facsimile images into tiles of various magnifications (for zooming) and for scaling the images.

The site consists of three main components: (1) the tiles, (2) the program that handles the queries to the index, and (3) the rest: static html pages, javascript files, stylesheets, and image files. Because of the difference in workload for each of these components, each is stored on a dedicated web server. For load balancing, the 3 servers are mirrored where necessary.

To hyperlink to individual letters and related documents, use links of the following form: https://vangoghletters.org/orig/let001 (to refer to the original-language version) or https://vangoghletters.org/en/let001 (to refer to the translation). Replace 'let001' by e.g. 'RM01' to create a link to the first related manuscript.

The source files for this edition are available here. In the zip file we also include the so-called 'ODD' file which is used to customise the TEI schema as well as the schema files generated from the ODD-file.5151. More about the process of TEI customisation: http://www.tei-c.org/Guidelines/Customization/. These files are available under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license.