Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
87 lines (70 sloc) 4.08 KB

Changelog for v1.7.0

This guide lists the most conspicuous change features for this release, and hopefully provides help for developers to achieve backwards compatibility.

Integration with Guzzle HTTP component

This guide cannot give a complete overview of Guzzle HTTP's awesomeness, but here some highlights:

  • Uploading files is now much quicker
    • Parallel requests using multi-cURL, which means files can be uploaded concurrently
    • Files over 5GB are fragmented and uploaded using multibyte chunking
  • URLs are now treated as fully fledged objects (Guzzle\Http\Url), allowing for greater functionality
  • Requests that send content now have the benefit of using the Guzzle\Http\EntityBody object, which extends and builds on the native functionality offered by PHP streams. Now there is a cool abstraction layer that offers a lot of consistency
  • Progress functionality (either by dedicated methods or by anonymous functions)
  • Event dispatching so all of your objects now have access to, and can also modify, the HTTP workflow. See Symfony's Event Dispatcher Component for more info.

For a complete list of features, view the official page.

Keypair support

Although code now exists in the SDK for API compatibility for creating servers with keypairs, there has been a recent bug in Nova where certain endpoints (DFW, IAD and ORD) do not exhibit the desired functionality. A patch will be released in the near future. For reference, you just need to add in an extra array key:

try {
        'name'     => 'test',
        'image'    => $image,
        'flavor'   => $flavor,
        'networks' => array(
        'OS-DCF:diskConfig' => 'AUTO',
        'keypair' => array(
            'name'      => '',
            'publicKey' => file_get_contents('/path/to/local/key')
} catch (\Guzzle\Http\Exception\ClientErrorResponseException $e) {

Better handling of services

  • Endpoints, the Service Catalog, and Catalog Items are now treated as objects
  • Factory methods in the core OpenStack client have been renamed for greater clarity
  • Services are now constructed with a ServiceBuilder for efficiency
  • Namespace changes to all resource models for greater consistency

Major upgrade of CloudFiles

  • Near 100% API consistency for Cloud Files
  • Complete documentation, both technical and non-technical, for all functionality
  • Better handling of resource objects: Account, Container, CDNContainer, DataObject
  • The ability to upload files in 3 different ways (each one has a different use case):
    • Sequential uploads for simple files
    • Parallel batched uploads for multiple files
    • An OpenCloud\ObjectStore\Upload\TransferBuilder factory for handling large uploads (+5GB) that require chunking
  • Conditional requests for getting objects (If-Match, If-None-Match, If-Modified-Since and If-Unmodified-Since). Range is also now supported for partial content returns.
  • Bulk extract and bulk delete now better supported

Unit test structure overhaul

  • Completely refactored unit tests
  • All unit tests now inherit from one centralized test case
  • Use of singletons for client object re-use
  • New implementation of mock responses
  • Request capturing using Event dispatchers

Other changes

  • Refactoring of Autoscale/Monitoring to keep things DRY
  • New docblock for every file with up-to-date license
  • Fix to logging output functionality
  • Fixes to server metadata
  • Minor changes to Metadata
  • Core change to method/property overloading. For every private model property, the end-user now has access to a dynamic getter/setter methods that ensure better encapsulation. For finer-grained control, they can specify concrete methods which supercede the magic methods. I want to get into the habit of encouraging filtered access through methods rather than raw property access.