From: Enoch <ix...@ho...> - 2013-04-02 19:26:53
|
Matthias Trute <mt...@we...> writes: >> Maybe it is a trivial observation, but I see that 2/ is a bad idea for >> addresses > 32KWhatever (bytes or words) since sign propagation will >> produce an incorrect address. I defined u2/ myself for that purpose. > > Its not the address but the size information. The chances are too little to > hit this bug (you need plenty of external RAM for this). Hello Matthias, What I expressed support for was adding u2/ to the standard, as recommended by Rafael. People with strictly typed language experience may easily forget about this subtle point. Regards, Enoch. |