Now the situation is every time, we will do the enhancement/fixing in the releasing branch for all the "approved" features, every round about impacting 2000+ files, with 5000 check-ins. We need to finally merge these code with the main trunk, then conducting code review, then build and test.
However, during above procedures, especially after we have done above. There would be some changes on CIs, selective features, and/or new "Hot" features. It means what we done has been wasted, and we have to start from another trunk to do the same.
Finally we will be failed to deliver timely, and team morale will be affected, so defects happened during the manual code merge.
I need experts to help us, how to have a better approach to deal with above situations? Or, any good practices can we have to make the auto code merge to reduce the defects?
Thank you very much in advance.
--- Sincerely yours, hhw.