elasticsearch versions numbering algorithm

216 views Asked by At

In the versions history of the Elasticsearch one can see a nontrivial suquence of versions, e.g.

Elasticsearch 1.4.2
December 16, 2014
Elasticsearch 1.3.7
December 16, 2014
Elasticsearch 1.4.1
November 26, 2014
Elasticsearch 1.3.6
November 26, 2014
Elasticsearch 1.4.0
November 05, 2014
Elasticsearch 1.3.5
November 05, 2014
Elasticsearch 1.4.0.Beta1
October 01, 2014
Elasticsearch 1.3.4
September 30, 2014

I would appreciate if anybody can explain (or give a link to appropriate document) why such versioning is being applied.

What is the reason to develop 1.3.x when 1.4.x versions are already available?

When 1.4.x are being developed, 1.3.x further development will consist only in bug fixes, or new features are being added also?

1

There are 1 answers

1
Dennis On BEST ANSWER

Assuming Elasticsearch uses Semantic Versioning, they are backporting patches for maintenance reasons (e.g. security vulnerability). This bumps the PATCH version (the "x" in 1.3.x).

No new features would be added to 1.3.x. Whenever a new feature is added, the MAJOR or MINOR version is bumped (e.g. 1.3.x → 1.4.x).

Disclaimer: I'm not an Elasticsearch maintainer so I can't say with certainty what versioning system they use.