[cfe-commits] r150055 - in /cfe/trunk: include/clang/Basic/DiagnosticGroups.td include/clang/Basic/DiagnosticSemaKinds.td test/Sema/switch.c

[David Blaikie]

On Wed, Feb 8, 2012 at 1:02 PM, Ted Kremenek <kremenek at apple.com> wrote:
> On Feb 8, 2012, at 10:17 AM, David Blaikie wrote:
>
> I'm also reminded of Chandler's talk at GoingNative last week when he
> mentioned that the -Wparentheses warning didn't used to have the
> inverse (checking that you /don't/ use extra () when you're just doing
> an equality not an assignment) & this was added. But I know it's not a
> perfect analogy because the extra () were already
> deliberately/habitually for -Wparentheses in the first place. Good
> that we could have negative & positive warnings on at the same time by
> default in that case.
>
>
> A fair point.  I think it's worth pointing out that a fair number of people
> disliked this change, so much that they requested that we could put the
> inverse case under a different warning flag so they could turn it off.
>
> I'm leaning that this should be under -Wall after all, but it would be great
> to hear other opinions here.

I don't have a clear idea either way (as I don't have a lot of
experience supporting compilers for a large number of teams/projects)
& I think you're right - knowing how much pain this is causing
existing users is important (the other half of the puzzle - knowing
how much value it'll give new (& existing - if they choose not to
suppress it) users is harder to quantify).

- David