220 4425 article Path: news.gmane.org!not-for-mail From: Fernando Cacciola Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Re: Fixing Standard Library relational operators (was: Re: optional Rev.4 (N3672): What was the rationale to remove !=, >=, ...) Date: Wed, 15 May 2013 09:20:32 -0300 Lines: 192 Approved: news@gmane.org Message-ID: References: Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: multipart/alternative; boundary=047d7b3441681edd2004dcc0c93b X-Trace: ger.gmane.org 1368620475 32661 80.91.229.3 (15 May 2013 12:21:15 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Wed, 15 May 2013 12:21:15 +0000 (UTC) To: std-proposals@isocpp.org Original-X-From: std-proposals+bncBCOPV7FAXQPRBOX3ZWGAKGQELZINXPQ@isocpp.org Wed May 15 14:21:16 2013 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-lb0-f198.google.com ([209.85.217.198]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1Ucahn-0000pa-UQ for gclcip-std-proposals@m.gmane.org; Wed, 15 May 2013 14:21:15 +0200 Original-Received: by mail-lb0-f198.google.com with SMTP id r10sf1574424lbi.9 for ; Wed, 15 May 2013 05:21:15 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=x-received:x-beenthere:x-received:received-spf:x-received :mime-version:in-reply-to:references:from:date:message-id:subject:to :x-original-sender:x-original-authentication-results:reply-to :precedence:mailing-list:list-id:x-google-group-id:list-post :list-help:list-archive:list-subscribe:list-unsubscribe:content-type; bh=LkHvlAfD5JAs1M3NUMh2Pf8MdjFtUUE0T1OFHc7qPlk=; b=R5cEIoIWNpNbEBKAv5f5U7nwIFCpj8I8X+EF4HKBi1glKzMRe4FaexOCnyh32zS93q ULKlMopcDCVzoEcmBxflHnTjCKWe5ICWKwbS8n+aDp02OxkymwDZesEN92qCda9J0Cf0 IFrGn1yESn3KVST92hZPGaroG4FmhauPcUFhc+1jC27oUquN4xqgkYRDmMlKqR7T+jhn YHpf88T5sKjTKdA5Hf6QXjDLthEQjsq7YHMCLFWtWNwfW8zcGAcbGR4NabymYcrNxO3B mqh+LEjKn2F2+WNgrae5Ss3O6C1r6BrYWE4EDQ4rfrYJzmmjRJm8iBGczEQqJ4p1WUVm X-Received: by 10.152.26.38 with SMTP id i6mr10847972lag.9.1368620475266; Wed, 15 May 2013 05:21:15 -0700 (PDT) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.152.27.230 with SMTP id w6ls49468lag.68.gmail; Wed, 15 May 2013 05:21:13 -0700 (PDT) X-Received: by 10.152.88.40 with SMTP id bd8mr18004975lab.9.1368620473622; Wed, 15 May 2013 05:21:13 -0700 (PDT) Original-Received: from mail-la0-x22b.google.com (mail-la0-x22b.google.com [2a00:1450:4010:c03::22b]) by mx.google.com with ESMTPS id nr2si990598lbb.9.2013.05.15.05.21.13 for (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Wed, 15 May 2013 05:21:13 -0700 (PDT) Received-SPF: pass (google.com: domain of fernando.cacciola@gmail.com designates 2a00:1450:4010:c03::22b as permitted sender) client-ip=2a00:1450:4010:c03::22b; Original-Received: by mail-la0-f43.google.com with SMTP id ez20so479101lab.2 for ; Wed, 15 May 2013 05:21:13 -0700 (PDT) X-Received: by 10.112.147.170 with SMTP id tl10mr17183362lbb.100.1368620472915; Wed, 15 May 2013 05:21:12 -0700 (PDT) Original-Received: by 10.114.80.164 with HTTP; Wed, 15 May 2013 05:20:32 -0700 (PDT) In-Reply-To: X-Original-Sender: fernando.cacciola@gmail.com X-Original-Authentication-Results: mx.google.com; spf=pass (google.com: domain of fernando.cacciola@gmail.com designates 2a00:1450:4010:c03::22b as permitted sender) smtp.mail=fernando.cacciola@gmail.com; dkim=pass header.i=@gmail.com Precedence: list Mailing-list: list std-proposals@isocpp.org; contact std-proposals+owners@isocpp.org List-ID: X-Google-Group-Id: 399137483710 List-Post: , List-Help: , List-Archive: List-Subscribe: , List-Unsubscribe: , Xref: news.gmane.org gmane.comp.lang.c++.isocpp.proposals:4425 Archived-At: --047d7b3441681edd2004dcc0c93b Content-Type: text/plain; charset=ISO-8859-1 On Wed, May 15, 2013 at 5:48 AM, Ville Voutilainen < ville.voutilainen@gmail.com> wrote: > > > > On 8 May 2013 01:29, Fernando Cacciola wrote: > >> On Tue, May 7, 2013 at 7:14 PM, Lawrence Crowl wrote: >> >>> >>> The naming is unfortunate, but the distinction between value ordering >>> and representation ordering is fundamental and deserves explicit >>> recognition in the language. Right now, we have the convention that >>> value ordering is spelled operator< and representation ordering is >>> spelled std::less. >>> >>> Well, I was not aware that this was, or is, the intention, but I do >> agree with the distinction between the two types of orderings, specially in >> the case where there is no value ordering at all, but there (always) is >> some valid representation ordering for the sake of lookups. I often resort >> to constructing a (large enough) integer out of the bits of all fields just >> for that. >> >> I also very much like the idea of using < and std::less for one and the >> other. >> >> I also agree that, for the sake of associative lookup, defininng >> std::greater in terms of std::less is exactly right, regardless of how >> operator > would be defined (so I'm saying I've got it wrong all along >> thinking that > in terms of < is a good idea because that's how the std >> functors work) >> >> Having said all that, I now feel that the appropriate resolution for >> std::optional<> is to implement all operators in terms of the underlying >> type. The rest of the library should IMO do the same, but if it can't be >> changed, let's at least avoid perpetuating the mistake. >> >> >> > I did a little bit of further homework on the matter. boost::tuple does > NOT synthesize the comparison operators, > but rather calls the operator of the underlying type. The existing > practice isn't consistent. ;) > > This needs at least an overview paper. I'd be willing to entertain the > idea to propose making all relational > operators, including tuple and containers, to not synthesize. I suppose > the fallback plans are synthesize-for-all > and deliberately-inconsistent, not necessarily in any preference order. > > In addition to that, tuples and containers would seemingly need std::less > specializations. That's actually somewhat separate. > > Fernando, Tony, would you be interested in coauthoring? > Now that I am much more convinced that the right approach is to not synthesize (consistently across the whole library), I would definitely like to coauthor such a paper! Best -- Fernando Cacciola SciSoft Consulting, Founder http://www.scisoft-consulting.com -- --- You received this message because you are subscribed to the Google Groups "ISO C++ Standard - Future Proposals" group. To unsubscribe from this group and stop receiving emails from it, send an email to std-proposals+unsubscribe@isocpp.org. To post to this group, send email to std-proposals@isocpp.org. Visit this group at http://groups.google.com/a/isocpp.org/group/std-proposals/?hl=en. --047d7b3441681edd2004dcc0c93b Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable
On W= ed, May 15, 2013 at 5:48 AM, Ville Voutilainen <ville.voutilaine= n@gmail.com> wrote:



On 8 May 2013= 01:29, Fernando Cacciola <fernando.cacciola@gmail.com> wrote:
=
On Tue, May 7, 2013 at 7:14 PM, Lawrence Cr= owl <crowl@googlers.com> wrote:

The naming is unfortunate, but the disti= nction between value ordering
and representation ordering is fundamental and deserves explicit
recognition in the language. =A0Right now, we have the convention that
value ordering is spelled operator< and representation ordering is
spelled std::less.

Well, I was not aware that this was, or is, the= intention, but I do agree with the distinction between the two types of or= derings, specially in the case where there is no value ordering at all, but= there (always) is some valid representation ordering for the sake of looku= ps. I often resort to constructing a (large enough) integer out of the bits= of all fields just for that.

I also very much like the idea of using < and std::less f= or one and the other.

I also agree that, for the sake of = associative lookup, defininng std::greater in terms of std::less is exactly= right, regardless of how operator > would be defined (so I'm saying= I've got it wrong all along thinking that > in terms of < is a g= ood idea because that's how the std functors work)

Having said all that, I now feel that the appropr= iate resolution for std::optional<> is to implement all operators in = terms of the underlying type. The rest of the library should IMO do the sam= e, but if it can't be changed, let's at least avoid perpetuating th= e mistake.



I did a little bit of further homework on the matter. boost::t= uple does NOT synthesize the comparison operators,
but rather calls the = operator of the underlying type. The existing practice isn't consistent= .. ;)

This needs at least an overview paper. I'd be willing to= entertain the idea to propose making all relational
operators, includin= g tuple and containers, to not synthesize. I suppose the fallback plans are= synthesize-for-all
and deliberately-inconsistent, not necessarily in any preference order.
=
In addition to that, tuples and containers would seemingly = need std::less specializations. That's actually somewhat separate.

Fernando, Tony, would you be intereste= d in coauthoring?

Now that = I am much more convinced that the right approach is to not synthesize (cons= istently across the whole library), I would definitely like to coauthor suc= h a paper!


Best


--
Fernando Ca= cciola
SciSoft Consulting, Founder
http://www.scisoft-consulting.com

--
 
---
You received this message because you are subscribed to the Google Groups &= quot;ISO C++ Standard - Future Proposals" group.
To unsubscribe from this group and stop receiving emails from it, send an e= mail to std-proposals+unsubscribe@isocpp.org.
To post to this group, send email to std-proposals@isocpp.org.
Visit this group at http://groups.google.com/a/isocpp.org/group/std-pro= posals/?hl=3Den.
 
 
--047d7b3441681edd2004dcc0c93b-- .