Re: Weird byte-compiler tests

Hrvoje> I've always considered the marker arithmetics a relic from ancient Hrvoje> times when someone thought it oh-so-cool to be able to increment a Hrvoje> marker or compare it to a number or another marker without bothering Hrvoje> to call `marker-position'. The marker arithmetics is important for Hrvoje> backward compatibility, but in this case we can just as well document Hrvoje> that the result types should not be relied on; they are numerically Hrvoje> correct, but their type cannot be known in advance and can depend on Hrvoje> whether the code is interpreted or compiled. I agree. On the other hand, it is hard to convert (char-before x) to (char-after (1- x)) if there weren't any marker arithmetics at all, without adding type checks.