[cfe-dev] Warning on strange indent

[Philip Reames]

On 02/25/2014 12:33 AM, David Chisnall wrote:
> On 23 Feb 2014, at 23:58, Nico Weber <thakis at chromium.org> wrote:
>
>> motivated by the current libsecurity_ssl brouhaha
> The one that would have been trivially avoided by a coding convention that said that if statements must always have their bodies enclosed in braces?  A rule that's been recommended by CERT and others for a very long time?  A rule that also avoids the dangling else problem?  A rule that is also the opposite of what LLVM recommends?  A rule whose lack has already caused subtle bugs in LLVM (I've fixed a couple of them in the past, I don't know how many other people have encountered)?
>
> David
> Who thinks coding styles should be designed to discourage bugs, not encourage them.
Yes, and I even agree with you.  Having said that, legacy code doesn't 
always comply and I'm not yet willing to run clang-format at that kind 
of scale.  Or get into the argument about style frankly.

Philip