Glossary
Due to changes in the Codex of Chess Composition 2009 this article is outdated; but on the other hand it may still be valid for problems prior to these changes. See also Castling and En-passant capture.
In some situations where a retrograde analysis proves mutual dependency (or mutual exclusion) between several elements pertaining to legal future moves, it is sometimes interesting to consider that the position is in fact a set of twin positions with different histories entailing different potential futures. They are retro-variants.
In that case, it is customary to insert "(RV)" in the stipulation. This draws the solver's attention to the fact that the several different possibilities have to be considered (and solved) in turn, usually with different solutions.
Here is an example:
W. Langstaff The Chess Amateur, 1922
5+3. (RV) Mate in 2
Here if Black still has the right to castle, then his last move was 0 ... g7-g5. Following the RV convention, we consider the two possibilities:
Notice that under the usual convention the problem is not solvable: White is not allowed to capture en passant and Black is allowed to castle ! Also, the RV viewpoint opposes the priority rule for mutually exclusive castlings.
Such situations led several authors to argue that RV should be the default viewpoint when mutual dependencies occur. However, the logic behind it is not perfect. E.g. in the example we just saw, it could be argued that there is a third possibility: Black can't castle and the last move was g7-g5. And here both 1. hg6 and 1. Ke6 mate in two. Clearly this is not the intention of the author.
One sometimes meet (PRA) (for Partial Retro-Analysis) instead of (RV). This was the earlier way of stipulating that the retro-variants had to be considered in turn.