reorder docs

README.md

@@ -11,10 +11,11 @@ Formats
 -------
 
 * [CrossRef API](http://api.crossref.org/), works and members
+* JATS [Journal Archiving and Interchange Tag Set](http://jats.nlm.nih.gov/archiving/versions.html)
 * [OVID](http://rzblx4.uni-regensburg.de/ezeitdata/admin/ezb_export_ovid_v01.xsd) holdings
-* FINC [Intermediate Format](https://github.com/miku/span/blob/ca8583aaa9b6d5e42b758f25ade8ed3e85532841/finc/schema.go#L67)
+* [Google holdings](http://scholar.google.com/intl/en/scholar/libraries.html)
+* FINC [Intermediate Format](https://github.com/miku/span/blob/master/schema/README.md)
 * FINC [SOLR Schema](https://github.com/miku/span/blob/ca8583aaa9b6d5e42b758f25ade8ed3e85532841/finc/solr.go#L4)
-* JATS [Journal Archiving and Interchange Tag Set](http://jats.nlm.nih.gov/archiving/versions.html)
 
 A little toolkit
 ----------------

schema/README.md

@@ -1,5 +1,28 @@
-Schema
-======
+Intermediate Schema Specification
+=================================
+
+The intermediate schema serves, among others, these purposes:
+
+1. It should assists clients in generating OpenURLs.
+
+2. It should assists clients in generating citation formats.
+
+3. It should provide an intermediary between data sources and export formats.
+   Instead of implementing m x n transformations for m data sources and n export formats, it reduces the effort to m + n.
+
+4. It can serve as a catch-all format, leveling out peculiarities of input data formats.
+
+The default serialization is JSON. Documents can be validated against a JSON schema (draft 4). The schema is versioned.
+
+Minor updates shall not break clients. Major updates may break clients.
+
+Notes
+-----
+
+* Languages can refer to the language of the article or the abstract.
+  Articles and abstracts both might use more than one language. The [ISO639-3](http://www.sil.org/iso639-3/) language codes are used.
+
+----
 
 The `is-<version>.json` file contains the JSON schema.
 Versioned examples can be found under the `fixtures` directory.