OpenSearch
Unfortunately, OpenSearch is two completely different things. Or, actually three. The Open Search Foundation is a European-based group trying to create a transparent, free Internet Search.
Open Standard[edit]
The OpenSearch specification (Wikipedia article) - an open standard for publishing search results came first in 2005. Ironically, it was developed by an Amazon subsidiary[1]. OpenSearch was (and still is) a way for people to directly use search engines from their browser. In addition to web search results, users could get product search results from Amazon, encyclopedia results from Wikipedia, movie information from IMDb, and countless more. Famously, MediaWiki includes this capability in the software, which means that you can directly search any wiki site (not just Wikipedia) from your browser.
There are many references to the OpenSearch standard (local article) on this site - explaining how to implement and use this search feature.
Software[edit]
Almost 20 years later, OpenSearch (software) is a family of software products created by Amazon in 2021 as a fork of Elasticsearch due to a licensing (money) dispute. The project is organized under the umbrella of the OpenSearch Software Foundation which is a project of The Linux Foundation organized to support the OpenSearch software project. The mission of the OpenSearch Software Foundation is to provide infrastructure and other resources to enable the long-term sustainability of the OpenSearch open source project and the OpenSearch ecosystem. Leadership starts at $150K annually. It is spearheaded by contributions from Amazon, SAP and Uber.
OpenSearch is a Lucene-based search engine that started as a fork of version 7.10.2 of the Elasticsearch service.
OpenSearch Dashboards started as a fork of version 7.10.2 of Elastic's Kibana software, and is also under the Apache License, version 2.
Version 3[edit]
Version 3.0 has been released as of May 6, 2025[2] - based on Lucene v10 and Java 21 plus support for the Java Platform Module System.
Experimental support for protocol buffers (protobuf[3]) over gRPC transport
Experimental Model Context Protocol (MCP) support on both the server and client side as an experimental feature, enabling seamless integration with AI agents.
Convert from Elasticsearch 7.x to OpenSearch 2.x[edit]
Since this site uses Elasticsearch for its search backend and also as the backend for Semantic MediaWiki, we are following the WMF transition to use OpenSearch as a replacement. This is tracked in Issue #133 on GitHub.
References[edit]
- ↑ A9 (short for Algorithm) The chief product was a browser toolbar.
- ↑ Blog announcement https://opensearch.org/blog/unveiling-opensearch-3-0/ Release Notes https://github.com/opensearch-project/opensearch-build/blob/main/release-notes/opensearch-release-notes-3.0.0.md
- ↑ Protocol Buffers is a free and open-source cross-platform data format used to serialize structured data. Think JSON, but faster (and less readable).
