Hi, I've seen recently some development with a special development flow that is the source of my feature request. Description of the feature: For each treated file, in case of an empty ancestor (file created on both branches or orphan branch), I found that usually the best way to help (or even do) the merge is to override the empty ancestor/base by the most recent between the oldest creation of the file of each branch. It could be: - an improvement of recursive strategy - new strategy that would be recursive with the special empty ancestor case - a new option for recursive strategy indicating how to treat empty ancestor
exemple: orphan branch1 created 2 weeks ago, develop branch existing since years. File a.txt created 2 years ago on develop but only 2 weeks ago on branch1. In such case, it's very likely that it has been created from a "git checkout develop -- a.txt" command. So taking that creation as the ancestor will propose a very good ancestor to look for possible merge process That's some things I've seen on projects that for some reason want to have only the minimum of files on each feature branch. It can happen usually when what's inside git (the source code) is not actually code that can be built but metadata describing the configuration of a product that's used to deploy the parameters. Thanks for reading this request. Kevin GOSSENT This message contains information that may be privileged or confidential and is the property of the Capgemini Group. It is intended only for the person to whom it is addressed. If you are not the intended recipient, you are not authorized to read, print, retain, copy, disseminate, distribute, or use this message or any part thereof. If you receive this message in error, please notify the sender immediately and delete all copies of this message.
