# Changelog

This project uses [Semantic Versioning](http://semver.org).

## 3.2.0

* Delete files in parallel

## 3.1.2

* VCR is not threadsafe. Because of this, disable parallelism during tests.

## 3.1.1

* Pass the file reference to each thread

## 3.1.0

* Upload files in parallel

## 3.0.0

* Set `text/html; charset=utf-8` as the `content_type` for each `text/html`
  object.

  This change may cause unexpected behaviour for some users. As a consequence,
  bump up major version.

## 2.8.0

* Add support for simple redirects

## 2.7.0

* Allow s3_id and s3_secret to be omitted

  This enables deploying jekyll-s3 on EC2 instances with IAM roles. The AWS SDK
  will automatically fetch temporary credentials.

## 2.6.0

* Support configuring of redirects on the S3 website

  This is possible thanks to version 1.2.0 of
  [configure-s3-website](https://github.com/laurilehmijoki/configure-s3-website),
  which is a dependency of the `jekyll-s3` gem.

## 2.5.1

* Respect the most specific `max_age` setting (fixes issue
  [43](https://github.com/laurilehmijoki/jekyll-s3/issues/43))

## 2.5.0

* Use regex to ignore files on server

## 2.4.3

* Make `s3_endpoint` optional in `_jekyll_s3.yml` when using `jekyll-s3` as a
  library
* Load dotfiles also with Ruby 2.0.0

  The `Dir.glob` implementation changed a bit in Ruby 2, and this resulted in
  `jekyll-s3` not uploading files that start with a dot.

## 2.4.2

* Fix problem where gzipped files were always re-uploaded

  See the issue
  [#29](https://github.com/laurilehmijoki/jekyll-s3/issues/29) for more info.

## 2.4.1

* Declare internal methods private in `upload.rb`

  This eases future refactorings

## 2.4.0

* Add support for specifying `Content-Encoding: gzip` HTTP header
* Add support for specifying `Cache-Control: max-age` HTTP header

## 2.3.0

* Add support for non-standard AWS regions

  You can now define the region with the `s3_endpoint` in the configuration
  file.

## 2.2.4

* Print the web site URL instead of the bucket URL
  ([#22](https://github.com/laurilehmijoki/jekyll-s3/issues/22)).

## 2.2.3

* Allow bugfixes for transitive dependencies *filey-diff*,
  *simple-cloudfront-invalidator* and *configure-s3-website*, since they are
  semantically versioned

## 2.2.2

* Remove debug code that wrote the list of site files into /tmp/test.txt

## 2.2.1

* Upload also dot files
  ([#16](https://github.com/laurilehmijoki/jekyll-s3/pull/16))

## 2.2.0

* Automatically configure the S3 bucket to function as a website.

  This makes it easier to start hosting a Jekyll site on S3. It removes the need
  to manually configure the S3 bucket in the Amazon AWS console.

## 2.1.2

* Remove a superfluous comma from uploader.rb. The comma might have caused
  problems for some Ruby 1.8.7 users.

## 2.1.1

* Remove optional settings from the generated `_jekyll_s3.yml` file

## 2.1.0

* Added support for S3 reduced redundancy storage

## 2.0.0

* Set content type of the uploaded files
* (Increment major version because of the changed return value in
  Jekyll::S3::CLI::run)

## 1.0.0

* CloudFront invalidation on changed files only
* Start using [Semantic versioning](http://semver.org/)

## 0.0.7

* Headless mode

## 0.0.6

* Upload only new or changed files
* Support ERB syntax in `_jekyll_s3.yml`

## 0.0.5

* Invalidate the Cloudfront distribution of the Jekyll S3 bucket.