Adam Fedor schrieb: > > On Jan 26, 2007, at 7:48 AM, Richard Frith-Macdonald wrote: > >> >> On 24 Jan 2007, at 16:36, Ted Howard wrote: >> Perhaps a bigger issue is that my compiler (gcc-4.1.1) happily >> compiles the current code! Now that seems to me like a rather severe >> compiler bug. Does anyone know different? > > I don't know why it wouldn't be a bug. Anyway, the Testfarm shows > several machines that fail (and don't). For instance gcc 4.0.1 on > Darwin fails to compile, but 4.1.1. (on Solaris) happiliy compiles this... > > http://www.gnustep.org/developers/testfarm-results.txt >
Sorry for taking so long to fix this simple bug. The ivar in NSMenuItemCell was totally useless and was only set and used within a single method. But then the code did not belong into this method at all. So I had to move it to a different method and then one thing followed from the other. In the end I had do do quite a few changes to the file and would even prefer to change more in it. But before doing this, I would like to get some feedback on the changes up to now. As for the compiler, I now think that the correct behaviour was to allow the duplicate ivar. In NSMenuItemCell this ivar was declared as private. With that scope it should not clash with another ivar of the same name in a super class. At least this is what I think. Still it is better to avoid this case and have unique names for ivars. Cheers, Fred _______________________________________________ Gnustep-dev mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnustep-dev
