Adopting the Schedule-Driven Release Process for Thunderbird

[Mark Banner]

The Firefox team is currently moving to a schedule driven release process, this has various implications for Thunderbird development.

There are two options that the Thunderbird drivers have considered:

• Adopt a new process that is schedule driven and follows the Firefox process of being able to release every six weeks (even if we don't release every six weeks).
• The advantages here are that we (have the option to) keep users up to date with the current gecko versions and security releases.
• It'll also mean we can get out more significant improvements to the users on a cycle which the current security and stability releases don't allow.
• We recognize that there are disadvantages, too. For instance, binary extensions will need to be rebuilt for each release.
• In their document, Firefox have highlighted some of the issues around extensions and we will be working on picking up what Firefox is doing to alleviate those issues.
• Maintain the current Thunderbird release model and maintain old releases for a period of time after the current release.
• The significant issue here is that back-porting security fixes from Firefox will require extra work, which doesn't help to advance Thunderbird. Back-porting may be additionally complex if the original patch writers are not available to help us.

The Thunderbird drivers have decided to adopt the new schedule driven process. We believe that it will be better overall for Thunderbird development and for our users.

We realise that this change may not suit all enterprise situations, however we feel that we need to move forward with Thunderbird development, and back-porting would slow development down. We suggest that enterprises use the tb-enterprise email list for discussion if this plan will affect them.

We are currently thinking about how to manage our release processes wrt the new Firefox branches and hope to have more information out in the next few days.

Mark