On Dec 6, 2012, at 11:28 PM, Дмитрий Дьяченко <dimhen@xxxxxxxxx> wrote:
> FYI
> gcc-4.8 will be more useful with -Wshadow :
> http://gcc.gnu.org/ml/gcc-patches/2012-10/msg02473.html
>
> may be, use -Wshadow only with new compiler?
Or "with a compiler that doesn't worry about variables shadowing functions", i.e. strengthen the test we're currently using for -Wshadow to check for that as well.
However, we've already fixed some instances of "index" as a variable (although perhaps it's time for UN*Xes still offering index() and rindex() to deprecate or just drop that, to encourage people to go with the Shiny New 1989 ANSI C Standard version, which really dates back to at least System III in the early 1980's), so -Wshadow doesn't currently have an issue there. If it pops up again, we could strengthen the configure-script test rather than changing code - or we could revert those changes and strengthen the test.