Helios M4 + 1 is next Monday December 14 so this Friday's (December 11) build will be our M4 candidate. I've already updated our build to build against Eclipse 3.6 so this will be our first milestone build against Eclipse 3.6.
When can we expect API breaking changes to be in HEAD? Can we postponed them till after M4 so consumers have more time to adapt to the new changes? Thoughts?
IBM Eclipse CDT
IBM Canada Toronto Lab
Doug Schaefer ---12/02/2009 11:26:38 AM---Hey gang, At the CDT call, we had a good discussion about moving CDT to 7.0 for the
At the CDT call, we had a good discussion about moving CDT to 7.0 for the Helios train in June. There are a few APIs that need to be changed to allow us to make the changes we need to do this release. There is also a strong desire to have firm and stable APIs and it would be good not to change them. That conflicts making this a tough decision.
But here's my thinking. I'm very glad we have the desire to have stable APIs and am very glad we are committed to managing them with the API tooling. However, until we have good APIs, I'm not sure we benefit from stable APIs. I don't think we're ready to commit to the APIs that we have as the set we want to stick with for the foreseeable future.
So basically, this means major releases until we get there. Unless we get a veto, Helios will be CDT 7.0. Committers need to commit to minimizing the API changes they do make to ensure the community can easily adopt CDT 7. And they need to commit to documenting the changes so that we can produce an upgrade guide. I think that's the most pragmatic approach.
I also think next year will be CDT 8.0. I encourage all contributors to take a look at the APIs we have and make the changes necessary to commit to them. I know the Build APIs are in much need of cleaning up. We also need to decide what exactly the API should be and remove things that we don't want. But the hope is that CDT 8.x is the last major release for a while.