It seems to me, like the parent said, he's just simplifying a normal process. Sure, if his code was a few lines, and someone deleted a few you could use version control to get them back. On the other hand, if his code was a full fledged program, and the deletions were institutional such that further work was carried on after the deletions making the entire datflow go through different assumptions and processes, it would be significantly more difficult to reinstate. Preserving the dataflow might be done if you're just commenting out a line to test things. But if a whole chunk is straight up deleted, the flow of data will likely shift to be incompatible with the old way of doing things within a couple of revisions of the code such that no version control would help.