Re: glib's configure.in and G_DISABLE_CAST_CHECKS

[Tim Janik <timj gtk org>]

On Mon, 4 Nov 2002, Owen Taylor wrote:

> It looks like like for GLib itself, we never turn on
> G_DISABLE_CAST_CHECKS, without regard to the --enable-debug
> setting. Isn't something like the following change to
> configure.in desirable?
>
> [ It's sort of trivial since the number of cast checks
>   in libgobject is tiny, but.. ]
>
> Regards,
>                                         Owen
>
> @@ -222,8 +222,10 @@ if test "x$enable_debug" = "xyes"; then
>
>    GLIB_DEBUG_FLAGS="-DG_ENABLE_DEBUG"
>  else
> +  GLIB_DEBUG_FLAGS="-DG_DISABLE_CAST_CHECKS"
> +

looks fine to me.

>    if test "x$enable_debug" = "xno"; then
> -    GLIB_DEBUG_FLAGS="-DG_DISABLE_ASSERT -DG_DISABLE_CHECKS"
> +    GLIB_DEBUG_FLAGS="$GLIB_DEBUG_FLAGS -DG_DISABLE_ASSERT -DG_DISABLE_CHECKS"

shudder, i still don't understand why someone would want to develop
against glib with G_DISABLE_CHECKS, or run it with G_DISABLE_ASSERT.
are people actually using --enable-debug=no?

>    fi
>  fi

---
ciaoTJ