Changes

Changes that affect behaviour or expected functionality will be listed here. This does not list all commits - refer to git log for that.

vNext

• Renamed a couple of internal setting names that had never been documented. In theory, this shouldn't break anyone but if you were relying on undocumented behaviour then be aware...
  • stalled_threshold has become stalled_threshold_days
  • colors/stalled has become stalled_color
  • colors/blocked has become blocked_color
• Added settings: options to the example projects so you can customize settings.
• Added the setting 'extra_css_filename' so you can customize the CSS of the generated report. Note that this doesn't help with the charts themselves as they aren't customized via CSS.

v2.0.1 (April 23, 2024)

• Fixed exception in rare case where history has an entry with no items
• Improved error messages when an issue cannot be parsed. The issue key will now be included in the output.

v2.0 (April 22, 2024)

Warning

Things that had previously been deprecated but had continued to work, have now been removed.

If you've been paying attention to deprecation warnings up to now then there should be no surprises in here. However, if you've been ignoring warnings then this version is a potentially breaking change. See also the page on deprecated warnings

• [Removed] Quite a while ago, we changed the directory where issues would be loaded from but the logic still remained to look in the old location as well as the new. Everyone should be on the new by now so removed that unneeded code.
• [Removed] The old way of configuring grouping rules was deprecated over two years ago but still worked. No longer supported.
• [Removed] Board ids being specified on the download block had been deprecated and are no longer supported.
• [Removed] Project keys, JQL and filter ids being specified right on the download block had been deprecated and are now removed.
• [Removed] AgingWorkTable: passing the priority_name right into the chart had been deprecated and is now removed
• [Removed] Chart total_wip_over_time_chart had been deprecated and replaced with daily_wip_by_age_chart. Now it's removed.
• [Removed] Chart blocked_stalled_chart had been deprecated and replaced with daily_wip_by_blocked_stalled_chart. Now it's removed
• [Removed] Passing the priority_name directly into the expedited chart has been deprecated for two years. Specify that in the board declaration.
• [Removed] Explicitly calling the discarded_changes_report was deprecated because it's always generated now.
• [Removed] In the Issue class, the method status_id was deprecated in favour of status.id. Removed.
• [Removed] status_category_mapping used to take a type parameter but that was deprecated because we don't need it anymore. Removed now.
• [Removed] The story point estimation chart used to support a grouping but that was deprecated when we changed the type of the chart. Removed.
• [Removed] If you still had some really old data with the old format for statuses, it was continuing to load because we had redundant code just for that. Removed that extra code.
• Generally tightened up all the logic around loading statuses, to account for the edge cases introduced by team-managed projects. There will be cases where you'll need to explicitly add a project id to the config and you'll do that as shown. The id won't be needed in most cases so I'd leave it off until the tool complains about needing it.

project id: 5 do
  ...
end

• Aggregated project sample:
  • Added a WIP chart by parent.
  • Added a dependency chart that only shows dependencies that cross boards
• Remove null fields from the JSON that came back from Jira, before saving them locally. On some instances, this simplistic compression can reduce the JSON file size by half.

v1.5 (April 16, 2024)

• Turns out that the board information returned by Jira Data Centre does not contain the project id and this was causing an exception. Fixed.

v1.3 (Apr 5, 2024)

• Added partial support for NextGen (team managed) projects in order to fix bug #22 where a NextGen project anywhere in the instance could cause jirametrics to blow up, in rare cases. The bug looks like this even though you hadn't set a status_category_mapping:

/jirametrics/project_config.rb:205:in `add_possible_status': Redefining status category Status(name="FakeBacklog", id=10012, category_name="To Do", category_id=2) with Status(name="FakeBacklog", id=10017, category_name="In Progress", category_id=4). Was one set in the config? (RuntimeError)

• Tweaked the standard_project a bit

v1.2.1 (Oct 5, 2023)

• Fixed some misleading text on the AgingWorkBarChart
• Fixed bug with loading the English gem that appeared to only happen on linux systems

v1.2 (Sep 14, 2023)

• Fixed bug in link creation when Jira isn't hosted at the top level of the domain.

v1.1 (Aug 22, 2023)

• Added the command 'go' that will do a download, followed by an export
• Fixes a bug when Jira chooses not to return a creator element in the issue. Theoretically impossible but it's been seen in production.

v1.0.2 (Aug 3, 2023)

• Added --name parameter to export and download so that we can selectively work with a subset of the named projects. Wildcards are supported just as in a terminal. Usage: jirametrics export --name "a*" to export only the projects starting with 'a'.
• Fixed bug when including individual projects into an aggregated one. If issues had been excluded in the individual project, they were still appearing in the aggregated.

v1.0.1 (July 3, 2023)

• Commands are now through the jirametrics tool, rather than rake. So you will execute jirametrics export rather than rake export
• Project is now packaged as a gem 'jirametrics'. This changelog will now list changes by version number instead of by date. gem install jirametrics
• Project has been renamed to jirametrics from jira-export because there was already a ruby gem by that name.

June 2023

• Added completed items to the Daily WIP by Blocked and Stalled chart. In the process, fixed a bug that had been causing miscounts in the total.
• Fixed bug where an item could be incorrectly marked as Blocked by Status in some cases where it was actually stalled

May 2023

• Issue.blocked_percentage has been removed. We're guessing that you didn't even know it was here and won't care that it's gone. The method was incomplete, in that it didn't track all possible ways an issue could be blocked, and generally shouldn't have been there in the first place.
• Certain statuses can be set up as 'stalled' and will reflect as stalled on all the reports. Configuration details are with project_config

April 2023

• StoryPointAccuracy chart now has the ability to use other types of estimates such as TShirt Sizes
• StoryPointAccuracy chart now uses size, rather than colour to indicate how many data points are at a specific spot. Also, it provides the option to see aging work in addition to completed.
• A side effect of above, is that now the grouping tag has been disabled as it doesn't make sense anymore.
• AgingWorkInProgressChart will now display an extra [Unmapped Statuses] column if there are any aging issues that aren't actually visible on the board. It's clearly a mistake if we have aging issues that aren't visible and this will make that more obvious.
• DailyWipByBlockedStalledChart now shows how long something has been stalled or why it's blocked.
• Removed some methods on Issue. stalled_on_date?, blocked_by_flag?, in_blocked_status_on_date? All of these are replaced with the new logic in Issue.blocked_stalled_changes() which provides far more detailed information about each state.
• The issue class now includes links in reasons that an item could be blocked.
• AgingWorkBarChart now uses times rather than dates when plotting blocked and stalled, which results in more accurate data.
• AgingWorkBarChart now displays in hover text why it's blocked. If it's blocked on a status then you'll see the status. If it's blocked on other issues, you'll see the issue ids.
• AgingWorkBarChart stalled now shows as starting right away rather than not starting for a full five days, which had been misleading in the past.
• Fixed bug where Issue.first_time_in_or_right_of_column would fail after anonymization
• AgingWorkBarChart now shows why an item is blocked (Flagged, status, etc) and not just that it is blocked.
• Certain statuses can be set up as 'blocked' and will reflect as blocked on all the reports. Configuration details are with project_config

March 2023

• AgingWorkBarChart will now grow the chart if there are too many issues. Previously, the chart would omit issues when too many were there and that was very misleading. It was also very hard to read when there were large numbers of issues.
• Fixed exception in the rare case where a status that is no longer accessible to the project is still mapped to 'backlog' on a Kanban board. This is another "theoretically impossible" situation that we've now seen in a production system.

January 2023

• Fixed bug where we didn't properly identify the backlog statuses in some cases
• Fixed bug where the log file name wasn't always being displayed during download.
• The log now writes to the root, rather than the target directory. In configurations that would switch target directories mid-download, it was ambiguous where it would actually show up.
• Explicitly throw exceptions if status names can't be found. The positive is that this will catch typos in confiration files more easily. The negative is that it may require more status_category_mapping declarations
• Fixed bug where sprints weren't showing up on sprint burndown in rare cases
• Now supports the personal access key form of authentication. Why does Jira have two different kinds of access tokens?

December 2022

• Aging work table now shows when the work was expedited.
• Fixed bug in aging work table where some blocked and stalled times weren't showing properly.
• Expedited priority names are now specified on the board, not on the individual charts. Also, you can now specify multiple different names.
• Added the canvas method on all charts so that you can override the default dimensions of any chart. Rarely needed when visualizing a single teams data but becomes valuable when visualizing multiple teams at once.
• Fixed a bug where the data could be messed up if some groupings didn't have a colour set.
• Cleaned up the output. What goes to $stdout is now readable and the full messy log goes to a file
• Aging Work Burndown now shows blocked and stalled.

November 2022

• Fixed bug where the final line segment wasn't being drawn for an active sprint in the burndown

October 2022

• project_key, filter, and jql declarations under the download section no longer work. See the deprecated page for details. Specifying a board id is now enough to pull all the right issues for that board.
• Many changes under the covers to prepare for aggregated projects, where we pull a bunch of projects together to show a common report across all of them.
• Fixed bug where items that were already expedited on creation, might not show on the expedited report.

September 2022

• Moved all quality checks to their own section at the top of the page. This will make the report more compact
• Moved discarded report into normal data quality checks
• Added a new quality check that will identify if subtasks are starting while the main issue is not.
• The 'items moving backwards' quality check no longer reports on items that were moved to the backlog. Those are covered in a different report
• The daily WIP charts have been completely rewritten to make them easier to extend and customize. More details to come on how you can customize them in your own reports.
• The two daily WIP charts (by age and by blocked/status) now show items that have completed but never satisfied the criteria for started. These are problematic so we're making them more visible.
• Some of the methods to configure reports have been renamed to be more consistent. If you use the old name then you'll get a deprecated warning along with the new name to use. In general, the names of these reports have been very inconsistent so you'll see more of these consistency changes as we go forward.
• Colours and ranges are now configurable on the story point estimate chart
• Fixed an interesting bug in the downloader where Jira would occasionally not download some issues that were in a sprint but had not yet been started. Only affected Scrum boards. After updating to the latest code, the code will do one full download before resuming incrementals and this is to ensure that we've downloaded all the issues we might have missed.
• Added summary statistics to the scrum burndown chart.

August 2022

• Pagination wasn't working properly when there were multiple pages of sprint data. Fixed.
• More quality checks for the story point accuracy chart.
• Added :backlog option when checking to see if an item was moved back to the backlog
• Added fix versions to the aging table. Also put an indicator when the current status isn't visible on the board. The really old work in the aging table is sometimes there because the issue is no longer visible on the kanban/scrum board and the team has therefore forgotten about it. These changes will make it more obvious when this is the case.
• If you don't specify any of project_key, filter, or jql then we look at the configuration for the board (assuming you set exactly one) and will use the filter specified there.
• The anonymizer would periodically blow up due to a bug in RandomWord. Put in retry logic to work around that.
• Fixed bug where a start and stop that happened at the same time could put the stop before the start. This made several charts inaccurate - particularly the WIP charts.
• Put in a quality check to look for issues that stop before they start. There are several cases that could cause this (not just the bug above) and when it happens, the charts become a mess.

July 2022

• Added status column to aging work table
• All charts can now customize the header and description by using header_text and description_text in their config
• All charts can now customize which quality checks are shown. The most common use case will be to turn all checks off but you can now tweak it any way you want. Use check_data_quality_for with no parameters to disable them all.
• Issue.stalled_on_date? will now look across subtasks (if present) to see if there was activity on those.
• Downloader now relies on the status category of 'Done' rather than checking the resolved field. Turns out that it's not mandatory to use resolved and for some projects that meant the downloader was pulling down massive amounts of data every time.

May 2022

• Issue.status now returns a Status object rather than a string.
• Dependency Chart: Labels for issues can now be overridden in rules
• Comments now get included in changes which means that the stalled report will now take comments into account
• Cycletime Scatterplot, Cycletime Histogram, and Aging Work in Progress charts now all support the same grouping rules that are in Throughput Chart.
• Throughput Chart now defaults to splitting data by issue type and also displaying a total line. It's still configurable with grouping rules if you want the old behaviour.

April 2022

• Throughput Chart now supports grouping rules
• Speed improvement for incremental downloads.
• Fixed exception when timezone_offset wasn't specified
• New chart: Dependency Chart to show all the links between issues. This requires that graphviz be installed. See docs.
• New chart: Story Point Accuracy to compare story point estimates to actual measured cycle times.
• Fixed bug where both burndowns were always being displayed regardless of which option was specified.
• Table headers are now "sticky" which means they'll stay on screen as you scroll. Requires newer browsers.
• Fixed bug in the cycletime scatterplot where the 85% lines could be incorrectly placed.
• Fixed bug where the html_report was requiring a board id even in cases where it wasn't ambiguous.
• jql has been reintroduced as an option on the download element so you aren't limited to project_name and filter_name.
• A debug message has been added to the downloader to make it more obvious how many more issues there are to pull.
• When we're pulling configuration information, if the board is "scrum" then we also download sprint data.

March 2022

• Changed the API we use to collect statuses so in theory, you won't have to add status_categories_mappings to the config anymore.
• Support for incremental downloads so we're only pulling those items that have changed, rather than all items in the system.
• Changed how issues are stored in the target directory. We now store them as individual files rather than in the groupings that JIRA had originally returned. Done to support incremental downloads.
• Added discard_changes_before so we can pretend that items moved back to the backlog hadn't actually started. If you do use then then also add the discarded_changes_report so you can see how this impacted any metrics you generate. Allowing items to move back to the backlog after being started is a poor practice as it hides the reality of what's been done. Think carefully before allowing that or using this feature.
• Changed quality notes to use tables instead of bullets for improved readability
• Improved the error messages in the case where we can't determine what status-category a particular status belongs to.
• Started tracking changes here