From fde1a74d7692fb993b71f52cf42932adab7b9d37 Mon Sep 17 00:00:00 2001 From: Weimin Ouyang Date: Fri, 5 Aug 2016 10:30:18 -0700 Subject: [PATCH] Update README.md Fixed some MongoDB term misuses. However, the whole description may need a careful inspection to make sure everything is consistent. --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index ebcbe3d..5097bf5 100644 --- a/README.md +++ b/README.md @@ -18,7 +18,7 @@ The data storage system should ideally be a database, rather than a file storage ##Technology Overview The current proposal recommends [JSON Schema](http://json-schema.org/) for modelling manifests and the [MongoDB](https://www.mongodb.org/) NoSQL database for storage. Web forms can be generated from and validated against JSON schema, and the resulting data can be formatted for storage in MongoDB's JSON-like format. Content can be parsed easily into human-readable YAML or parsed directly by processing scripts for automated processes. -MongoDB is very well suited to these considerations. A MongoDB record is a single JSON object called a “collection”. Each collection consists of a set of keyword-value pairs called “documents”. These terms are confusing in the context of WE1S, so, in the discussion below, I will use the term “record” for MongoDB’s “collection”, and I will refer to keywords and their values, or sometimes “fields”, rather than “documents”. A MongoDB record looks like the following: +MongoDB is very well suited to these considerations. A MongoDB record is stored as a single BSON object called "document". Each document consists of a set of keyword-value pairs. These terms are confusing in the context of WE1S, so, in the discussion below, I will use the term “record” for MongoDB’s “document”. A MongoDB record looks like the following: ###Example 1: @@ -123,4 +123,4 @@ Some initial evaluation of Alpaca has been undertaken. Although these technologies are primarily recommended for managing research workflow, a goal of the project is to make data and provenance queryable by the public. In general, MongoDB is not the best system for complex data queries because it lacks the database joins of which most relational databases are capable. How much of a problem this would be depends on the data and the type of queries you expect to run. Aggregating data in the application’s code, as rather than in the database query, can have an impact on performance, but in most cases it is possible to achieve the same result. There is a body of thought that a document storage system like MongoDB can be a stepping stone to eventually move the data into a relational database with a more rigid schema. This is something that can only be assessed based on usage tests. -[Continue to the Manifest Schema](https://github.com/scottkleinman/WE1S/blob/master/DraftSchema.md) \ No newline at end of file +[Continue to the Manifest Schema](https://github.com/scottkleinman/WE1S/blob/master/DraftSchema.md)