@branneman branneman released this May 23, 2016

  • Feature: Renamed finished-when to wait-for-script (dbb2e5f)
  • Feature: Now allowing wait-for-script on global, page and component level (#12, 7643770)
  • Feature: Added wait-for-delay option, which will idle for a specified amount of milliseconds (c93c8a4)
  • Feature: Added run-script option, which will run a script on the page enabling clicking stuff, opening of dialogs, etc. (f970952)
  • Feature: wait-for-script and run-script are now strings referring to javascript files, no more inline scripts (0130a9d)
  • Feature: Added credentials option to be used as HTTP Basic auth (ecaa664)
  • Feature: Added phantomjs-flags option (bf40f68)
  • Fix: Improved error message for ignore rule failures (b60fa34)
  • Chore: Updated internal dependencies
  • Chore: CI servers are now also testing for Node.js v6.x
  • Docs: Improved texts on the differences between wait-for-script and run-script


@branneman branneman released this Apr 8, 2016 · 24 commits to master since this release

  • Renamed add action to capture, (using add still works for now, but it's deprecated)
  • Default threshold is now 0
  • Refactored out the external ImageMagick dependency, now using BlinkDiff
  • Added timestamping to the verbose logging level
  • Performance: Now creating just 1 PhantomJS instance per page/url, instead of per 1 instance per component. (#11)
  • Performance: Now running PhantomJS in parallel, also added concurrency as an argument (6b385c5, #11)
  • Performance: now getting image size without loading it the entire image into memory
  • Tests: Replaced express with http-server (#8)
  • Fix: It's now possible to use a threshold of exactly 0 (#2)
  • Fix: Mapping between left and right fullpaths are now safe (1b61ab5)
  • Fix: Non-fatal errors in PhantomJS are now allowed (84fc386)
  • Fix: identifiers are now slugified, supporting all characters (d350452)


@branneman branneman released this Feb 10, 2016 · 56 commits to master since this release

  • Added sizes option to the config, specifying the screen size used in the captures [docs]
  • Added finished-when option to the config, to indicate when a page has finished loading, defaults to the DOMContentLoaded event [docs]
  • Added configtest action, which only validates a config file [docs]
  • Added a timeout of 10 seconds to the capture script [source]
  • Added functional tests [source]
  • Fix: It's now possible to run --help or --version without a (valid) config file