220 32082 article Path: news.gmane.org!.POSTED!not-for-mail From: "dgutson ." Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Re: Re: Non-last default arguments Date: Tue, 11 Apr 2017 23:43:41 -0300 Lines: 213 Approved: news@gmane.org Message-ID: References: <91d734b5-d1f2-773a-b5c2-3b96696a6f0f@scylladb.com> <23242c11-a20b-9dfe-42fc-7a1ae3195849@scylladb.com> <58EBA69B.2020902@gmail.com> <58ECE695.8060601@gmail.com> Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: blaine.gmane.org Mime-Version: 1.0 Content-Type: multipart/alternative; boundary=001a1147e8c42d3949054cef2d8b X-Trace: blaine.gmane.org 1491965025 11673 195.159.176.226 (12 Apr 2017 02:43:45 GMT) X-Complaints-To: usenet@blaine.gmane.org NNTP-Posting-Date: Wed, 12 Apr 2017 02:43:45 +0000 (UTC) To: std-proposals Original-X-From: std-proposals+bncBDE3NBMV6UFBBX5IW3DQKGQEJRI6KGY@isocpp.org Wed Apr 12 04:43:40 2017 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-yw0-f200.google.com ([209.85.161.200]) by blaine.gmane.org with esmtp (Exim 4.84_2) (envelope-from ) id 1cy8Ft-0002uf-T7 for gclcip-std-proposals@m.gmane.org; Wed, 12 Apr 2017 04:43:38 +0200 Original-Received: by mail-yw0-f200.google.com with SMTP id s39sf6288665ywa.7 for ; Tue, 11 Apr 2017 19:43:44 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=isocpp-org.20150623.gappssmtp.com; s=20150623; h=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-spam-checked-in-group:list-post :list-help:list-archive:list-subscribe:list-unsubscribe; bh=6OTDiESv1YhHZiBLscYKDy1fn0VU4jEtACfO6jl1ARs=; b=Svr8TmoLBIQv6VQBnSVSWOmTHCTx5EjmdU5GE+6UR02UIf4wcLU12UE/ZnCub82xa8 jevYB/QYZSeT3eTFxmiPIvEQrO115Hol3NAUBAysuJ8u35wsWAPOhLNO2Te+RTqjTGpv zjIwJAQyDBW22RGrKLS1xcTgfR/pMxqidG0e0fo/hpRPokj9M2zEgNJyBWvsCD0+QzjV OWWeJP7j8rlNrUszfup17oydiX6b78323+I33/X1Yu7Hq9WhzPx7/rOOXQCHhEWCUwQL mIAlUR0kKBsk6lbH9re3o/z3+chtosNL0kmmI0k6L6ZRvDYaanewn6x1v16lLFKDKGww tTTQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state: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-spam-checked-in-group:list-post:list-help:list-archive :list-subscribe:list-unsubscribe; bh=6OTDiESv1YhHZiBLscYKDy1fn0VU4jEtACfO6jl1ARs=; b=UsDU5mwsJXzKXaJu88uTjbuGszPikdw5kJaBmAz2qc5YOJY6EmBXVFMIKuL7toZjad g/2edbbDsk16ZDDIp8Si4Mv7xsq5QHi9XP9h2HwPowdr3yrfa86DfwnQybMR7OCs+MaR MAs+WNqqqvZ+n/QpLlctpfINGnLCN0AV6QP/pL+fLxfBXn9FQmUhjZYjyMZvmv2Owz/K 8QXEpzWPbcwSBrxvYz6HQClg7Tz6bfLLytJcIYnU+S2rD7QRka1ttH64NfUpVoyhTlXW Ou+6odyVL4WG/rgCLbOlB1vcmI3pPon0QL8r3eZeeOlunVSVDcMOrgPzWLETAzpH6143 5dcg== X-Gm-Message-State: AN3rC/6kgBIt0VFCCk3m2u1MSw4SeIZ+Z/JsYYjfInLadcceuitQFT9gaEazxVt8QjN44Q== X-Received: by 10.129.160.12 with SMTP id x12mr7559179ywg.90.1491965023661; Tue, 11 Apr 2017 19:43:43 -0700 (PDT) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.157.56.220 with SMTP id k28ls8110940ote.7.gmail; Tue, 11 Apr 2017 19:43:42 -0700 (PDT) X-Received: by 10.37.194.5 with SMTP id s5mr38380105ybf.136.1491965022848; Tue, 11 Apr 2017 19:43:42 -0700 (PDT) Original-Received: from mail-yw0-x231.google.com (mail-yw0-x231.google.com. [2607:f8b0:4002:c05::231]) by mx.google.com with ESMTPS id u5si6272441ybf.310.2017.04.11.19.43.42 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Tue, 11 Apr 2017 19:43:42 -0700 (PDT) Received-SPF: pass (google.com: domain of danielgutson@gmail.com designates 2607:f8b0:4002:c05::231 as permitted sender) client-ip=2607:f8b0:4002:c05::231; Original-Received: by mail-yw0-x231.google.com with SMTP id j9so6465175ywj.3 for ; Tue, 11 Apr 2017 19:43:42 -0700 (PDT) X-Received: by 10.129.95.67 with SMTP id t64mr2186386ywb.18.1491965022377; Tue, 11 Apr 2017 19:43:42 -0700 (PDT) Original-Received: by 10.129.162.148 with HTTP; Tue, 11 Apr 2017 19:43:41 -0700 (PDT) Original-Received: by 10.129.162.148 with HTTP; Tue, 11 Apr 2017 19:43:41 -0700 (PDT) In-Reply-To: X-Original-Sender: danielgutson@gmail.com X-Original-Authentication-Results: mx.google.com; dkim=pass header.i=@gmail.com; spf=pass (google.com: domain of danielgutson@gmail.com designates 2607:f8b0:4002:c05::231 as permitted sender) smtp.mailfrom=danielgutson@gmail.com; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=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:32082 Archived-At: --001a1147e8c42d3949054cef2d8b Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable El 11/4/2017 12:57, "Nicol Bolas" escribi=C3=B3: On Tuesday, April 11, 2017 at 10:22:16 AM UTC-4, Matthew Woehlke wrote: > > On 2017-04-10 23:10, Arthur O'Dwyer wrote: > > On Monday, April 10, 2017 at 8:37:03 AM UTC-7, Matthew Woehlke wrote: > >> On 2017-04-07 19:09, Arthur O'Dwyer wrote: > >>> You might look into how Boost does it with boost::use_default. I > >>> haven't looked myself and can't promise anything, but I gather > >>> that several Boost libraries have something like this: > >>> > >>> inline struct use_default_t {} use_default; > >>> template struct defaulted { > >>> optional value; > >>> defaulted(T); > >>> defaulted(use_default_t); > >>> operator T() const; > >>> }; > >>> > >>> void f(defaulted a, defaulted b, defaulted c, > >>> defaulted d); > >> > >> Ugh; you've promoted changing the default from a SIC to a BIC. > > > > (I guess "SIC"=3D"syntactic interface change" and "BIC"=3D"binary inter= face > > change". Consider spelling these abbreviations out on first reference.) > > Anyone that has ever read qt-devel knows these. Alas, Google is > unusually unhelpful here. > > SIC =E2=86=92 Source Incompatible Change > BIC =E2=86=92 Binary Incompatible Change > > (There are also SC and BC =E2=86=92 Source/Binary Compatible.) > > Actually, I realize that's slightly wrong. Changing a "normal" default > means that the code will still run, but recompiling will change the > behavior, so while it's technically both SC and BC, it is (potentially) > behavior breaking. (If done carefully, however, it can be okay to make > such a change.) Changing a "boost style default" not only (maybe) > changes behavior when the program is recompiled, it *is* an actual BIC, > i.e. your program will crash until it is recompiled because the ABI > changed. > If default arguments are going to be made more complicated, then I would genuinely want them to become part of the function's ABI. The current situation, where they're just a bit of syntacitc sugar, is fine for the simplest of cases. But when you get function calls like `func(default, value, value2, default, value3)`, I think it's really important that if that interface changes, everyone who call the function with them will need to be recompiled. Quite off topic but still relevant and familiar: I do prefer to solve default template arguments than function arguments. Many times I found myself trying to see what type of allocator or whatever defsult template argument was used in order to specify the next one. The trivial useless case, *just* to be illustrative: unordered_map. That's why I think using named parameters via designated initilizers is really the way to go to get this behavior. --=20 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. To view this discussion on the web visit https://groups.google.com/a/ isocpp.org/d/msgid/std-proposals/a08da561-78d7-4dfe- 8b22-d03b9b933ebe%40isocpp.org .. --=20 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 e= mail to std-proposals+unsubscribe@isocpp.org. To post to this group, send email to std-proposals@isocpp.org. To view this discussion on the web visit https://groups.google.com/a/isocpp= ..org/d/msgid/std-proposals/CAFdMc-1qSOjSpsDkG0n64k%2B%3Dngj85cFPpqaYg5%3DKZ= mK1Yx5BQQ%40mail.gmail.com. --001a1147e8c42d3949054cef2d8b Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable


El 11/4/2017 12:57, "Nicol Bolas" <jmckesson@gmail.com> escribi=C3=B3:
On Tuesday, April 11, 2017 at 10:22:16 AM UTC-4, Matthew Woehlke wr= ote:
On 2017-04-10 23:10, Arthur O&#= 39;Dwyer wrote:
> On Monday, April 10, 2017 at 8:37:03 AM UTC-7, Matthew Woehlke wro= te:
>> On 2017-04-07 19:09, Arthur O'Dwyer wrote:=20
>>> You might look into how Boost does it with boost::use_defa= ult. I=20
>>> haven't looked myself and can't promise anything, = but I gather
>>> that several Boost libraries have something like this:
>>>
>>> inline struct use_default_t {} use_default;=20
>>> template<class T, T default_value> struct defaulted = {=20
>>> =C2=A0 optional<T> value;=20
>>> =C2=A0 defaulted(T);=20
>>> =C2=A0 defaulted(use_default_t);=20
>>> =C2=A0 operator T() const;=20
>>> };=20
>>>
>>> void f(defaulted<int, 1> a, defaulted<int, 2> = b, defaulted<int, 3> c,=20
>>> defaulted<int, 4> d);=20
>>
>> Ugh; you've promoted changing the default from a SIC to a = BIC.=20
>=20
> (I guess "SIC"=3D"syntactic interface change" = and "BIC"=3D"binary interface=20
> change". Consider spelling these abbreviations out on first r= eference.)

Anyone that has ever read qt-devel knows these. Alas, Google is
unusually unhelpful here.

SIC =E2=86=92 Source Incompatible Change
BIC =E2=86=92 Binary Incompatible Change

(There are also SC and BC =E2=86=92 Source/Binary Compatible.)

Actually, I realize that's slightly wrong. Changing a "normal&= quot; default
means that the code will still run, but recompiling will change the
behavior, so while it's technically both SC and BC, it is (potentia= lly)
behavior breaking. (If done carefully, however, it can be okay to make
such a change.) Changing a "boost style default" not only (ma= ybe)
changes behavior when the program is recompiled, it *is* an actual BIC,
i.e. your program will crash until it is recompiled because the ABI cha= nged.

If default arguments are going to be made m= ore complicated, then I would genuinely want them to become part of the fun= ction's ABI. The current situation, where they're just a bit of syn= tacitc sugar, is fine for the simplest of cases. But when you get function = calls like `func(default, value, value2, default, value3)`, I think it'= s really important that if that interface changes, everyone who call the fu= nction with them will need to be recompiled.

Quite off top= ic but still relevant and familiar: I do prefer to solve default template a= rguments than function arguments.
Many times I found= myself trying to see what type of allocator or whatever defsult template a= rgument was used in order to specify the next one.
T= he trivial useless case, *just* to be illustrative:
= =C2=A0 =C2=A0 unordered_map<Key, T, default, myAllocator>.



That's why I think using named parameters via d= esignated initilizers is really the way to go to get this behavior.

--
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.
To view this discussion on the web visit = https://groups.google.com/a/isocpp.org/d/msgid/std-proposals/a08d= a561-78d7-4dfe-8b22-d03b9b933ebe%40isocpp.org.

--
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-proposa= ls+unsubscribe@isocpp.org.
To post to this group, send email to std-proposals@isocpp.org.
To view this discussion on the web visit https://groups.google.com/a/isocpp.org/d/msgid/std-proposals/CAFdMc-1qSO= jSpsDkG0n64k%2B%3Dngj85cFPpqaYg5%3DKZmK1Yx5BQQ%40mail.gmail.com.
--001a1147e8c42d3949054cef2d8b-- .