this is part of a crusade to get rid of compilation warnings

Comment on attachment 456739 [details] [diff] [review] patch I don't see any reason for a change like this to require per-module review. rsmoa=me iff the NS_DEBUG_ASSIGN thingy itself gets r+ My $0.02 on that: what would be wrong with just doing #ifdef DEBUG foo = #endif bar(baz); in all these cases? I agree it's butt-ugly, but it makes it much clearer what I think this is trying to do (and if I have misunderstood what it's trying to do, that actually strengthens my point ;-) ). What I like least about this it that it's not intuitively obvious what exactly is going to happen in a non-debug build. OT1H NS_DEBUG macros don't evaluate their arguments, OTOH the rhs isn't actually an argument of the macro, so what happens? (That's a rhetorical question -- I'm sure it has a simple answer, but the point is that the question shouldn't arise).

I'm glad you finally homed in on what I suggested 4 weeks ago ;-)