Agile development uses rapid releases to deliver features early and often to users, enabling them to provide quick feedback about the software. We observed that relevant projects, such as Firefox and Chrome, adopt rapid release, but in some cases, the development duration takes longer than the release cycle. This work analyzed whether other projects adopting rapid release also present similar behavior. We analyzed an open-source corpus comprising 1,039 relevant open-source projects, with 13,102 rapid releases and 7,071 traditional releases (without counting the patches), totaling 20,173 releases with 3,167,563 commits. We compared rapid and traditional releases to understand changes regarding the development duration, the development start delay, and the percentage of percentage of commits in the release cycle. We discovered that, in most cases, the development duration of rapid releases is higher than its release cycle, meaning that most projects delivering rapid releases employ parallel release development. However, on average, the amount of work done in parallel has a small impact on the release development. Moreover, the development of rapid releases starts early and without delays. Traditional releases’ development is generally preceded by a pause, probably due to patch development and release stabilization. Therefore, our results suggest that projects that intend to adopt rapid release do not necessarily need to reduce the development duration. The projects may start the release development early and manage parallel release development to achieve a rapid release cycle. This paper was submitted to IEEE Transactions on Software Engineering in October 2023 and is under review.