We use SVN for our version control and have a trunk for minor changes and a branch for more significant changes and also for the devs to be able to work off the same code base.
We regularly do trunk merges so our branch code-base stays up to date. When we work on a given bug / issue we do so generally on the branch and then commit all files related to that change.
We then merge that specific revision from the branch into the trunk and publish it.
This works most of the time, with occasional issues with merges. However, am I correct in thinking that eventually this is going to cause problems if we continue to use SVN's automatic merging?
My reasoning is that, for example, file A on branch is regularly tweaked, edited and committed by developer A and this may or may not be part of bug / issue commit, they might just be showcasing functionality.
Developer B gets the latest update of File A, makes changes related to a bug fix and commits. He then merges with the trunk using SVN's automatic merge functionality, no conflicts are reported.
Is it not possible for changes made by developer A to have ended up in the trunk because of the automatic merge? Using a HTML example, could an extra div be added because it didn't conflict with existing code but was 'new'?