Christopher Faylor wrote: > Ok. Again, I don't want to worry about the use of __extension__. If I > am not going to worry about it and Corinna doesn't want to be worrying > about it then I don't see any reason to do it. It doesn't make much > sense to use it if the two principles aren't interested. > > Case closed.
I'm sorry I replied and caused you further irritation, I thought this was still open to discussion. Since it isn't, I shall move on. Before I submit my next patch, I was hoping we might further discuss the use of "a ?: c", since you mentioned it was a conscious decision. As I noted previously, I thought it could be a source of confusion since this behaviour is not documented and is not widely know. My supposition was confirmed when even a seasoned programmer like Gary Van Sickle was not aware of this behavior. I, myself, had to rummage through the GCC source code to be absolutely certain what the implict behavior was. I propose expanding all such instances to "a ? a : c" to improve clarity of intent and code readability. Thanks in advance. Cheers, Nicholas
