Log message:
ruby-jirametrics: update to 2.9.1. Changes:

- Fixed an exception when running the tool for the very first time.

Log message:
ruby-jirametrics: update to 2.9. Changes:

- Status/category mappings have been an ongoing pain point for a long
  time. When a status gets removed from Jira, all of a sudden the
  reports can't be generated until someone sets up a
  status_category_mapping and it isn't always obvious what the values
  should be. We are now much better at guessing the correct mapping and
  will use that guess where we can. You will see suggestions in the
  output, where we guessed and what values we used.
- Bug: Fixed an exception in the data quality report when a status can't
  be found at all.
- On StatusCollection, there used to be methods todo, in_progress, and
  done that would return the names of statuses that matched those
  categories. Since we now know that names aren't unique, those
  methods have just been removed. They hadn't been documented before
  and were never officially supported so we aren't going through a
  deprecation process.
- Bug: Fixed case where the quality report didn't always correctly warn
  about items that had been moved back to the backlog.
- Whenever a status or status-category is specified, we can do a better
  job of identifying typos.
- The methods first_time_in_status_category, still_in_status_category,
  and currently_in_status_category...
    - ...will now accept names ("Done"), name/ID pairs \ 
("Done:3"), or
      just ID's ("3"). If the name and ID are both specified then we
      will verify that they match and raise an error if they don't.
    - ...will also accept one of the following keys :new,
      :indeterminate, or :done to reflect the three categories. The code
      had previously worked only with names but those names will change
      in non-English installations of Jira so the keys are better.
        - Why is it called "indeterminate"? That's a question for
          Atlassian; we just use the key names that Jira uses
          internally.
- standard_project now uses the keys, which should make things easier
  for non-English instances.
- Added a quality check for the case where an issue is identified
  as being blocked by another issue and yet that other issue is
  already closed. This is clearly a mistake and the quality report
  will now show it.

Log message:
ruby-jirametrics: update to 2.8. Changes:

The main focus of this release is improving the experience with better
error messages, and more helpful suggestions.

- Bug: It wasn't possible to clear the description or title texts on
  individual charts.
- Added first_time_label_added for determining start or end times.
- Moved the generated timestamp of the report to the footer.
- Added the version number of JiraMetrics to the footer so there's
  a record of what version of the tool was used for this
  particular report.
- no_earlier_than previously only worked during the download stage and
  this proved to be confusing as there was no effect when just doing an
  export. It now works at both points and should be clearer as to what
  is actually happening.
- All the methods like first_time_in_status used to return a Time object
  and now return the ChangeItem that happened at that time. Unless
  you're writing your own methods, you likely won't even be aware that
  anything changed here. If you are, then you'll get a warning when
  returning a Time. The reason for this change is to give more context
  that will be useful in future reporting.
- Bug: We've discovered that Jira allows for multiple statuses with the
  same name and when this happens, it broke some of our existing logic.
  Fixes for this case have been added.
  - In standard_project, when specifying status_category_mappings, it's
    now possible to specify the ID at the same time as the name.
    "Review" will continue to specify just the name and \ 
"Review:4" will
    specify the name and also an ID of 4.
  - Status ID's are now mandatory and the tool will do it's best to
    guess the correct one, however it won't be able to guess in every
    case and you'll now get a error when it can't.
- Bug: some status category logic was failing in non-English instances
  of Jira because we were relying on category names, rather than keys.
  We think we've caught all the cases of this.

Log message:
ruby-jirametrics: update to 2.7.3. Changes:

- Fixed an exception when two scrum projects had overlapping names like
  myproject and really_myproject. Our matcher wasn't precise enough and
  was matching where it shouldn't have.

Log message:
ruby-jirametrics: update to 2.7.1. Changes:

- Removed duplication in the data quality report
- Improved wording
- Removed unneeded code
- Updated URL's to point at JiraMetrics.org

Log message:
ruby-jirametrics: update to 2.7. Changes:

- Fixed misleading usage information when running jirametrics with
  no options.
- Fixed deprecated warning that would display if an unknown option was
  given to jirametrics
- Added info command from terminal for debugging purposes. If you are
  trying to figure out why the tool has calculated something in one way
  then this will dump out information about that particular issue that
  might be helpful. Usage jirametrics info ABC-123 where ABC-123 is the
  issue you want information on.
- Experimental: Trying out a way to visualize flow efficiency. I'm not
  convinced this is the right way yet but I'm making it public so I can
  get feedback. This may change. Add flow_efficiency_scatterplot to your
  configuration. If you want to try this from standard_project then
  specify the option show_experimental_charts: true.
- Fix: discard_changes_before was always intended to work either as an
  option under project or under html_report but there were cases where
  it was broken in the former case.
- Fix: There are legitimate cases where the calculated start and end
  times for an issue are exactly the same. Consider
  in_or_right_of_column where Done is legitimately to the right of that.
  In this case, we now pretend that the issue never started, which in
  almost every case will be the expected result anyway. In the cases
  where that isn't true, our configuration is probably wrong anyway.
- Deprecated: CycletimeConfig.started_time(Issue) and
  CycletimeConfig.stopped_time(Issue), which are now replaced with a
  single method that returns both:
  CycletimeConfig.started_stopped_times(Issue). This was necessary to
  fix the bug above, where started and stopped are the same.
- Added another quality check. It now checks for subtasks that aren't
  closed when the main issue is closed. This likely indicates that the
  issue is still in progress, despite what the status says.

Log message:
ruby-jirametrics: update to 2.6. Changes:

- When a status can't be found, it now just dumps a warning, rather than
  blowing up. This has been a source of frustration
  for many people just starting out with the tool so we're trying to get
  some more reasonable default behaviour.
- There is a new `Issue.flow_efficiency_numbers` method that is a first
  step towards visualizing flow efficiency. Note that none of the charts
  currently make use of this.
- By default, we assume that `flagged` means blocked but not all teams use
  the flag for that purpose. Added new setting to change that behaviour:
  `settings['flagged_means_blocked'] = false`.

Log message:
ruby-jirametrics: update to 2.5. Changes:

* Better error messages in the case where
  `settings['customfield_parent_links']` does not point to the
  correct field.
* Fixed exceptions when loading a scrum board that has no sprints.
* Added `no_earlier_than` option in `download` block. This is helpful
  when there were some big changes in the way we worked and we want to
  ignore all data before a specific date. This can be used together with
  `rolling_date_count` or by itself.
* `standard_project` now accepts `rolling_date_count` and
  `no_earlier_than` options.
* `expedited_priority_names` has been deprecated on the board block. Now
  set it through `settings['expedited_priority_names']`
* Provided more reasonable defaults for `DependencyChart`. Everything is
  still configurable if you don't like the defaults.
  * Less noisy in terms of console output.
  * Now excludes any link where both ends are already closed.
  * Now excludes `cloned by` that everyone was manually excluding anyway.
  * Now properly uses the calculated `done` rather than looking at status category.
  * `standard_project` and `aggregated_projects` now build on the new defaults.
* More reasonable defaults in settings. As above, these can be
  overridden if you don't like the defaults.
  * `"blocked_link_text"` now defaults to `["is blocked by"]`
  * `expedited_priority_names` now defaults to `['Critical', 'Highest']`