What version are you running? 2.0.13. I've also reproduced the issue on ReviewBoard's demo instance (currently, 2.5 beta 2) What's the URL of the page containing the problem? http://demo.reviewboard.org/dashboard/ What steps will reproduce the problem? 1. Create a new empty review. 2. Have a single user account give the review two "ship it!"s 3. View the dashboard. 4. Note that the dashboard shows a "ship it" count of two despite only one person having reviewed the code. What is the expected output? What do you see instead? The dashboard should show the number of people who have given the review a ship it instead of the total number of ship its received. What operating system are you using? What browser? Linux. Firefox. Please provide any additional information below. A review might have multiple ship its from one person if the diff is updated significantly and someone that has already given the review a ship it gives it another after reading the updated diff. Other reviewers browsing their dashboard might choose to skip the review if they see that it has multiple ship its, even though there was only one person that reviewed it and gave it a ship it.
For now, this is intentional. Say I'm working with someone and have a Ship It, but I realize I need to make another change, so I upload a new diff. Seeing the Ship It counter go up on the dashboard is useful information, in that it lets me know at a glance that the new diff was also approved. We may want to rethink this, but only in the context of a revamped Ship It system.
This confuses a lot of people where I work too. @chipx86, in your example, we'd want the new version to cause the previous ship-its to clear and the new count to be 0. We have a "you need a ship it on your final version from two different people" policy, so sometimes you see one reviewer giving ship its on every version, and another reviewer with a concern that lasts through 5 or 6 revisions, and the "ship it" count shows 7 or 8 by the end. Maybe have a different count option, "distinct ship its on latest version" and/or "distinct ship its" (not sure if the original request was for the same use case as ours)
Maybe this will help you, too.