[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [lmi] Clang fixes
From: |
Vadim Zeitlin |
Subject: |
Re: [lmi] Clang fixes |
Date: |
Thu, 12 May 2016 18:14:51 +0200 |
On Thu, 12 May 2016 00:50:33 +0000 Greg Chicares <address@hidden> wrote:
GC> On 2016-03-26 00:27, Vadim Zeitlin wrote:
GC> > On Fri, 25 Mar 2016 18:26:17 +0000 Greg Chicares <address@hidden> wrote:
GC> [...]
GC> > GC> But I'm not convinced that progress_meter::display_mode_ is
fundamentally
GC> > GC> a wrong idea, or that it can never be useful, so I'm reluctant to
spend the
GC> > GC> time and effort to root it out and rewrite the documentation.
GC> >
GC> > I understand this and don't propose to do it.
GC> >
GC> > GC> Therefore, is there a way to tell Clang that its objection is noted,
GC> > GC> but we've decided not to take its advice?
GC> >
GC> > There are plenty of ways to do it, starting from the direct one of just
GC> > disabling this warning with a clang-specific pragma to pretending that we
GC> > do use it, as you propose, but I don't like doing this at all.
GC>
GC> Sorry, to me, it's like performing an appendectomy without medical need:
But surely we're talking about engineering here and not something as messy
and imprecise as biology? I certainly hope we can do better than just
letting our program code blindly evolve on its own, accumulating vestigial
remains as it does...
GC> But we need to move past this clang obstacle, so I committed the change
GC> I proposed earlier.
Thanks, I'll test it soon, but I'm pretty sure that it's going to work. I
still can't fathom that you don't see the wisdom of pruning completely
unnecessary code, but if you really, really want to keep it, could we just
add a comment saying that this field is currently unused and is kept only
for the potential future developments? At least this will help the next
person reading this code.
Regards,
VZ