Review Assistant moved to SQL Server

In this release, the main efforts were directed to moving the server-end portion of Review Assistant to Microsoft's SQL Server as a database engine. We've also made a number of improvements in a review workflow.

Why move to SQL Server?

For a long time, Review Assistant has been using SQLite as a database engine. This was a perfect solution for small teams: the server runs faster and requires almost no setup.

Lately, the number of big teams among our clients started to grow. This revealed the problem with performance that couldn't be solved if we continued using SQLite. The major problems were caused by concurrent work of a large number of users.

In the new release, we expect our customers to notice the improved performace of the Review Assistant server. Our tests revealed that in some cases the server started working 2-3 times faster.

Besides that, a number of problems with excessive memory usage were eliminated.

Review Assistant supports all editions of Microsoft SQL Server starting from SQL Server 2008.

SQL Server as a database engine

Other improvements

  • Any comment in a review can be marked as fixed.

    Previously, an author could mark only "defect"-flagged comments as "fixed". Now, an author can mark any comment as "fixed" to convey to a reviewer that the action for the comment is completed. This will allow the author and reviewer to track the progress without closing the comment.

  • In a simple review workflow, a reviewer can close a review after code is accepted.

    If a reviewer has no further comments on a code, they can accept author's changes and finish the review.

  • The "search-as-you-type" suggestion appears when adding users to a project.

    As you type the name of a user into a filter box, the smart filter suggests only the relevant names based on the characters you have entered.

  • On the Code Review Board, projects in the drop-down list are now sorted alphabetically.

    A small enhancement for those who have a long list of projects and frequently switch between them.

  • Working with a long list of comments in a review became faster.

    Another small but nice improvement is that now, when you scroll through a large set of comments, the list smoothly runs down or up without any delays.