Differences between the development modes of CVs and SVN

Solution 1: Set the daily scheduled daily build, lock the database during build, and tag the database after BVT is completed. Then, the file is obtained as a package based on the differences between the two labels. This solution is suitable for productization and regular release tasks. This solution is relatively simple. During the submission process, you can use the lock database method (with authorization to submit) or do not lock the database to view the organization. In my opinion, it is better to lock the database and grant the submitter permission to the person with the task. Solution 2: Build the version based on the file entered by the user. This solution requires the SVN to lock the database and authorize the submitted file to be submitted. It is not a problem that CVS does not lock the database. Cvs can use the CVS Update-R command to update files by specific version number. Therefore, you can roll back the versions of a file at will, or release files of different version numbers from the same branch to different update packages; the most common scenario is that a file version 1.2 activity = open, version 1.3 activity = close. You can activate and close the activity package at the same time. SVN supports updating by version number, but this will lead to confusion in the version number of the local SVN library. Therefore, the latest version is obtained by default, so that only one package can be published before the next package.