On Tue, 24 Oct 2023 at 12:25, Benjamin Lerer <ble...@apache.org> wrote:
> My biggest concern with the 5.1 suggestion is that it makes the review of > TCM far more complicated than it should be. Even if all TCM patches were > fully reviewed by committers that I fully trust, due to the patch size and > the impact of the changes it feels safer to me to have a second round of > review now than the dust is settling. Merging TCM and Accord in a 5.1 > branch and doing the review there will make things harder than it should. > To be clear, in a cassandra-5.1 branch we would be performing this second round of review post-merge, which as you say makes it more complicated. But if we do this in the cassandra-5.0 branch, this second round of review is then to be done pre-merge – blocking both tcm and accord from any preview release that users can start playing with (i.e. users that want to play with this will have to build it themselves from the feature branch). Can you provide an estimate on how long we need for this second round of review, best-case and worst case scenarios ?
