From owner-svn-src-head@FreeBSD.ORG  Sun Jul 13 14:27:38 2014
Return-Path: <owner-svn-src-head@FreeBSD.ORG>
Delivered-To: svn-src-head@freebsd.org
Received: by hub.freebsd.org (Postfix, from userid 1033)
 id 80260601; Sun, 13 Jul 2014 14:27:38 +0000 (UTC)
Date: Sun, 13 Jul 2014 14:27:38 +0000
From: Alexey Dokuchaev <danfe@FreeBSD.org>
To: Hans Ottevanger <hans@beastielabs.net>
Subject: Re: svn commit: r268566 - head/usr.bin/users
Message-ID: <20140713142738.GA25502@FreeBSD.org>
References: <201407120747.s6C7lpYE020200@svn.freebsd.org>
 <20140712195452.N3279@besplex.bde.org>
 <53C12C1D.8080101@fastmail.net>
 <20140712132705.GA45964@FreeBSD.org>
 <53C290BE.8050102@beastielabs.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <53C290BE.8050102@beastielabs.net>
User-Agent: Mutt/1.5.23 (2014-03-12)
Cc: src-committers@freebsd.org, svn-src-all@freebsd.org,
 David Chisnall <theraven@freebsd.org>, Bruce Simpson <bms@fastmail.net>,
 Bruce Evans <brde@optusnet.com.au>, svn-src-head@freebsd.org
X-BeenThere: svn-src-head@freebsd.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: SVN commit messages for the src tree for head/-current
 <svn-src-head.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/options/svn-src-head>,
 <mailto:svn-src-head-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/svn-src-head/>
List-Post: <mailto:svn-src-head@freebsd.org>
List-Help: <mailto:svn-src-head-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/svn-src-head>,
 <mailto:svn-src-head-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Sun, 13 Jul 2014 14:27:38 -0000

On Sun, Jul 13, 2014 at 03:59:26PM +0200, Hans Ottevanger wrote:
> On 07/12/14 15:27, Alexey Dokuchaev wrote:
> > Now the question is, how do we come out of this mess?  Shall we redo
> > it in C again?  Polish C++ version?  Merge it with w(1) as suggested?
> 
> More in general, in my experience C++ is not that suitable as a language
> for this kind of small utilities, where you do not need its abstraction
> facilities. This could be the reason why the original C version was as
> good as and (if polished a bit) even better (clearer) than the C++ version.

Very good point.  C++ is a nice language when you need to approach some
complex problem that you won't be able to grok with C if you want to reach
the same level of correctness (god bless destructors), maintability, code
conciseness, etc.  For a project of those magnitudes, resulting code sizes
of C vs. C++ implementation would likely be statistically negligible, so
no one will even mention them. ;-)

> Additionally, one of the properties of a code tree that influences
> quality is uniformity. Rewriting random small C utilities that work
> perfectly well in C++ breaks uniformity.
> 
> I can imagine that some selected utilities (or groups of utilities that
> share abstractions) that are difficult to maintain when written in C are
> rewritten in C++, but some discussion in current@ or hackers@ beforehand
> might be in order.

I fully agree.  Thanks for spelling it out Hans.

./danfe