Re: [PATCH 00/37] removal of some c++ keywords

From: Duy Nguyen <pclouds@gmail.com>
Date: Jan 30, 2018, 00:13 UTC
Message-ID: <CACsJy8Cm8HsopKisiJkwtPyfv-O6Ei3waew6CsKLXzKv7=TriQ@mail.gmail.com>
In-Reply-To: <20180129223728.30569-1-bmwill@google.com>

On Tue, Jan 30, 2018 at 5:36 AM, Brandon Williams <bmwill@google.com> wrote:

> A while back there was some discussion of getting our codebase into a state
> where we could use a c++ compiler if we wanted to (for various reason like
> leveraging c++ only analysis tools, etc.).  Johannes Sixt had a very large

I would be more convinced if I knew exactly what leverage we could have here (e.g. what tool, how many problems it caught...).

Show 5 quoted lines

> patch that achieved this but it wasn't in a state where it could be upstreamed.
> I took that idea and did some removals of c++ keywords (new, template, try,
> this, etc) but broke it up into several (well maybe more than several) patches.
> I don't believe I've captured all of them in this series but this is at least
> moving a step closer to being able to compile using a c++ compiler.

Is it simpler (though hacky) to just  do

#ifdef __cplusplus #define new not_new #define try really_try ...

somewhere in git-compat-util.h?

Do we use any C features that are incompatible with C++? (or do we not need to care?)

> I don't know if this is something the community still wants to move towards,
> but if this is something people are still interested in, and this series is
> wanted, then we can keep doing these sort of conversions in chunks slowly.

You're going to need to setup C++ build job on Travis or something to catch new C++ keywords from entering the code base as well if you move this to the end.

-- 
Duy