Skip to content

Latest commit

 

History

History
122 lines (112 loc) · 7.72 KB

CHANGELOG.md

File metadata and controls

122 lines (112 loc) · 7.72 KB

v0.4.0 (5-Apr-2018)

  • Updated code to create task without trigger. #25 (#Vasu1105)
  • Fix for execution time limit and weeks of month. #23 (#Vasu1105)
  • Added methods to get and set principal information of the task. #22 (#Nimesh-Msys)
  • Added methods to retrieve settings(all/Idle/Network) of current task. #21 (#Nimesh-Msys)
  • Refactored constants, moved predefined MSDN constns to another file. #20 (#Nimesh-Msys)
  • Added code to handle on idle trigger and on idle settings. #19 (#Vasu1105)
  • Fix for trigger at_logon and at_system_start. #18 (#Nimesh-Msys)
  • Added get_task and enabled? method. #17 (#Vasu1105)
  • Fix for undefined method 'weeks' error while updating week_of_month#16 (#Vasu1105)
  • Fix for handling days of month for trigger_type MONTHLYDATE. #15(#Vasu1105)
  • Fix for setting system user for scheduled task. #14 (#Vasu1105)
  • Fix for Wrong value is set for end_day, end_year, end_month it should be EndBoundary and not StartBoundary. #13(#Vasu1105)

v0.3.2 (18-Mar-2017)

  • Use the block form for FFI::MemoryPointer in the error message helper function. Thanks go to Ethan Brown for the suggestion.
  • Fixed a potential bug in the helper module, which was also renamed to help prevent any name collisions.
  • Added the win32-taskscheduler.rb file for convenience.
  • Added the configure_settings method.
  • Added the configure_registration_info method.
  • Added the description and description= aliases for comments.
  • Added the author and author= aliases for creator.
  • Some internal cleanup, moving common code to private methods.
  • Rakefile now assumes Rubygems 2.0 or later for tasks.
  • Gemspec cleanup, updated home page, removed old rubyforge_project reference.
  • This gem is now signed.

v0.3.1 (6-Jul-2014)

  • Added FFI as a dependency. Thanks go to Maxime Lapointe for the spot.
  • Some updates to the win_error helper method. Thanks go to Ethan J. Brown for the suggestions.

v0.3.0 (15-Jan-2014)

  • Rewritten to use Win32OLE instead of using wrapping the C API. Benefits include working on Windows Vista or later, and working with JRuby.
  • Modified the constructor to accept 3rd and 4th arguments. These indicate which folder to use, and whether or not to create it if it doesn't exist.
  • The TaskScheduler#save method is now no longer necessary. It is retained for backwards compatibility, but will raise a deprecation warning. In this version simply calling TaskScheduler#activate will implement the task.
  • Added support for the AT_SYSTEMSTART, AT_LOGON and ON_IDLE trigger types.
  • Now requires the structured_warnings gem.
  • Removed the doc directory and separate documentation file. Everything is inlined now. There's still an example under the "examples" directory, too.
  • Added test-unit, rake, and win32-security as development dependencies. These are needed for testing only.

v0.2.2(29-Feb-2012)

  • Moved some include statements inside the TaskScheduler class to avoid polluting Object. Thanks go to Josh Cooper for the spot and patch.
  • Minor formatting tweaks to silence 1.9 warnings.

v0.2.1 (8-Oct-2011)

  • Fixed a bug that would not allow task to run as SYSTEM. Thanks go to Josh cooper for the spot and patch.
  • Fixed a bug in new_work_item that would cause it to crash if you tried to create a work item that already existed. An error is now raised instead. Thanks go to Pete Higgins for the spot.
  • The set_trigger and trigger= methods now internally transform and validate the trigger hash in the same manner as new_work_item. Thanks again go to Pete Higgins.
  • Cleaned up the repo. The C source files have been removed from the main repository (and this gem). They are in a separate branch on github for anyone who misses them.
  • Refactored the Rakefile, removing tasks related to the old C source files, and added tasks for cleaning, building and installing a gem.
  • Updated the README file, eliminating references to anything that was only related to the older C version.

v0.2.0(19-Jun-2009)

  • Rewritten in pure Ruby!
  • The TaskScheduler::ONCE constant is now a valid trigger type. Thanks go to Uri Iurgel for the spot and patch.
  • Added the TaskScheduler#exists? method.
  • Added the TaskScheduler#tasks alias for the TaskScheduler#enum method.
  • The TaskScheduler#new_work_item method now accepts symbols as well as strings for hash keys, and ignores case. Also, the keys are now validated.
  • Renamed the example file and test file.
  • Added the 'example' Rake task.
  • Fixed some code in the README synopsis that was incorrect.
  • The TaskScheduler#save instance method now accepts an optional file name.
  • Most of the TaskScheduler setter methods now return the value specified instead of true.
  • Removed the RUN_ONLY_IF_DOCKED and RUN_IF_CONNECTED_TO_INTERNET constants. The MSDN docs say that they are unused.
  • Added more documentation. Much more rdoc friendly now.
  • Added many more tests.
  • Better type handling for bad arguments.
  • Added a Rakefile with tasks for building, installation and testing.
  • Added a gemspec.
  • Inlined the rdoc documentation.
  • Internal project reorganization and code cleanup.
  • Bug fix for the bitFieldToHumanDays() internal function.
  • Moved the 'examples' directory to the toplevel directory.
  • Made the CHANGES and README files rdoc friendly.
  • Minor updates to taskscheduler.h.
  • Now uses the newer allocation framework and replaced all instances of the deprecated STR2CSTR() function with StringValuePtr(). This means that, as of this release, Ruby 1.8.0 or later is required.
  • Modified the constructor to accept arguments. This is just some sugar for creating a new task item in one call instead of two.
  • The argument to trigger= now must be a hash. The same goes for the 'type' sub-hash.
  • Added the add_trigger() method. Actually, the C code for this method was already in place, I simply forgot to create a corresponding Ruby method for it.
  • Removed the create_trigger() method. This was really nothing more than an alias for trigger=(). I got confused somehow.
  • Test suite modified and many more tests added.
  • Documentation updates, including docs for a couple of methods that I had accidentally omitted previously.
  • Initial release