A few weeks ago Evan and I had a long discussion about where we wanted to take Adium over the next few months. Ultimately the discussion turned to our release schedule…
Adium, beyond 0.63:
Adium has moved to the stage in its life where new features are either massive undertakings or target a small number of users. A one week release schedule does not leave enough time for massive additions (such as the new contact list, meta contacts, a file transfer interface, etc), and minor features do little to warrant a new release and little towards encouraging users happy with their current version to upgrade.
So the obvious solution was to extend the releases to a longer schedule. We agreed upon releasing every other week instead, with the idea that this would give us more time to work on each release of Adium and result in bigger, more encouraging to upgrade changelogs.
But then we got to thinking… At this point in Adium’s life, what are the advantages of a weekly release schedule? One clear advantage is that users receive more frequent updates. However, we’ve found that recently a lot of users aren’t downloading every new release of Adium. When asked, they tell us that are happy with the release they have and see nothing in a new version worth the hassle of a download. To understand why this happens we need to see Adium updates from a typical user’s perspective.
For most users, upgrading software brings both excitement and fear; Excitement over what has been added to the newest release and fear over what will be discovered as unintentionally broken. If excitement does not outweigh fear a user will be less likely to upgrade.
What Adium needs is more excitement and less fear. We want every version of Adium to be exciting and contain at least one change or addition that makes the program more enjoyable to use for every user. We’d like Adium to be as bug free as possible and would love for each release to correct more bugs that it introduces. We desire Adium releases that are more fun for our users and less of a weekly routine. The best way to achieve all of this would be switching to a more traditional release schedule.
While a weekly release schedule is based on time, a traditional schedule is one based on goals. The Adium team would create several goals for each release (0.7, 0.8, 0.9) and a release would occur once those goals had been completed. This is great because goal-based releases would be less frequent, leaving more time for testing and lowering fear. Goal-based releases would also be larger, bringing more additions and enhancements and raising excitement! As an added bonus, a traditional release schedule would make it much easier to release minor updates for bugs, security, and connectivity issues.
What does this mean to you as a user? Simply a better, more exciting Adium.
See you all in 0.70 🙂
Misc Notes:
– Users looking to live on the edge with daily ‘releases’ are encouraged to build Adium from source. We’ve switched from CVS to SVN and our source no longer has a 5 hour delay, which means that you can experience changes instantly as they are made and announced in #adium.
– Moving away from weekly releases makes working on Adium a lot more relaxing and provides us with a chance to experience a traditional release cycle. Most of the developers are here for fun and experience, and this change should provide us with more of both.