In your mind, does tagging a module for 5.0 imply that the module is 100% ready for production, or only that it is functional to the point of not causing any fatal errors or breaking the overall system?
I guess I'm looking for guidelines on how far along a module should be before being tagged for 5.0, given that Drupal core itself is a moving target right now, and core changes might break a module that previously worked with HEAD.
You asked Dries, but here's my opinion... ;) A module without a 5.0 tag does not tell an end user anything. It could be bleeding edge, but it could also be an outdated module. There is no way for them to know. If you are making an effort to maintain your module, then you should definitely tag it. If an end-user downloads a module tagged as 5.0, but it does not work, then they know that they can expect an update in the future. If it's not tagged as HEAD, they are left in the dark. Late tagging IMO slow down the release of Drupal core 4.7.0 before: too many contributed modules were not tagged, causing too many people to stick to 4.6. In spite of us releasing several betas, it took until the final release candidate for people to really start testing the release and finding some easy to spot bugs. In short: tagging, yes please, as soon as possible. Steven Wittens