220 19895 <7066b166-9472-4ad6-ac3c-2db527ed7818@isocpp.org> article Path: news.gmane.org!not-for-mail From: Arthur Tchaikovsky Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Re: default arguments Date: Tue, 18 Aug 2015 09:37:54 -0700 (PDT) Lines: 235 Approved: news@gmane.org Message-ID: <7066b166-9472-4ad6-ac3c-2db527ed7818@isocpp.org> References: Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="----=_Part_438_480074727.1439915875035" X-Trace: ger.gmane.org 1439915880 3487 80.91.229.3 (18 Aug 2015 16:38:00 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Tue, 18 Aug 2015 16:38:00 +0000 (UTC) Cc: mwoehlke.floss@gmail.com To: ISO C++ Standard - Future Proposals Original-X-From: std-proposals+bncBDL6FI4NYYERBZF6ZWXAKGQELC4I7MY@isocpp.org Tue Aug 18 18:37:59 2015 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-qg0-f69.google.com ([209.85.192.69]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1ZRjte-0001Qy-3s for gclcip-std-proposals@m.gmane.org; Tue, 18 Aug 2015 18:37:58 +0200 Original-Received: by qgj62 with SMTP id 62sf185320338qgj.3 for ; Tue, 18 Aug 2015 09:37:57 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=date:from:to:cc:message-id:in-reply-to:references:subject :mime-version:content-type:x-original-sender:reply-to:precedence :mailing-list:list-id:x-spam-checked-in-group:list-post:list-help :list-archive:list-subscribe:list-unsubscribe; bh=2YGWsPSbevCMJANtgyZfWeMUfiWbdnUp+WzzFGflXi4=; b=c+qR5DV9wcDg2TNRB5L6ywg2UMVh5XFF7Z0eAgseQmVsW5lolz3T7pkH8ZtwYECFlw DoGi753MjQKcfUsupaYutVLSV/af4Ou85CrH9J372jI0ytKskBDoxeI8gaS55H58WXwa qiMou4/WfifX7S8xIQ65p/Ityeq+XSPdsNalKDGlfhYBYCMZxw0fda4xd3QRCXN7tr7q FWVmXWhMpePyJUE2jUZBUO2e1naHkJ+wZhJU8nf8Y8zzhx82KZlAd59rBW22sVq3b6cV AvvtVxq/2Lw4u3ftSWQhtuJRr55pnBmBULyTUjOwThASKGP9Ob4SyvcvKpEsdJI0hnCT D0ig== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:date:from:to:cc:message-id:in-reply-to :references:subject:mime-version:content-type:x-original-sender :reply-to:precedence:mailing-list:list-id:x-spam-checked-in-group :list-post:list-help:list-archive:list-subscribe:list-unsubscribe; bh=2YGWsPSbevCMJANtgyZfWeMUfiWbdnUp+WzzFGflXi4=; b=LnLzG5m1Mq8o4+KSL6lP7VWPQLrOYgGW/4zkLYjgyb8pwtgiVEb0hzx0orFXgn2/65 ab/2NSid0yJVTX2yOBHd6ob6MNaCpt6EuOmX/xprB0XtYi2esmVD4HKTnAvkNhRY4P9N CI/AJ85N5KLsQN7cFlk09O4PyRNbMqj89zZgTnd5U0fXIo5AtiTUWOD7hHFi/eVRn81i 0x53wtuzfXKT+LmZ7NyXG+4Vi8CA5w7wRtiHRE+YvKrGy6YIVAhskCsQF/zJDY1GJl8E +BXN3M/bD3nT7np7r0FQSTHfNLEKhhHvnCQWJitYgbaTM/eGHx15HLdjlqKncEqD5HSk CmzQ== X-Gm-Message-State: ALoCoQmArEDKy1qdwFYJbONuJzxBghOU/+kMlTvhyCpuc6T/E6RmQQFlWveyXteawqGmQqzCiuYw X-Received: by 10.13.207.1 with SMTP id r1mr6863649ywd.53.1439915877113; Tue, 18 Aug 2015 09:37:57 -0700 (PDT) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.140.41.167 with SMTP id z36ls3376320qgz.29.gmail; Tue, 18 Aug 2015 09:37:55 -0700 (PDT) X-Received: by 10.140.36.170 with SMTP id p39mr103729qgp.28.1439915875770; Tue, 18 Aug 2015 09:37:55 -0700 (PDT) In-Reply-To: X-Original-Sender: atch.cpp@gmail.com Precedence: list Mailing-list: list std-proposals@isocpp.org; contact std-proposals+owners@isocpp.org List-ID: X-Spam-Checked-In-Group: std-proposals@isocpp.org 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:19895 Archived-At: ------=_Part_438_480074727.1439915875035 Content-Type: multipart/alternative; boundary="----=_Part_439_656562220.1439915875035" ------=_Part_439_656562220.1439915875035 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable I dear to disagree with the statement that the following example is self=20 documenting and easier to read: auto c3ptr =3D new C{.timeout =3D 1};=20 How is this^^^ easier to read? And how is this self documenting? Only on=20 surface it seems that it is, both, but when you really think about it it is= =20 neither self documenting nor easy to read. Why? Because having only one=20 argument initialized in {}, that is .timeout, leaves us with question, is= =20 that the only argument to that class, or are there any other? If so, are=20 those arguments inited by default? If so, how many? Etc. Compared to this: auto c3ptr =3D new C(default, 1); // use a shorter timeout=20 It is immediately obvious how many arguments there are and which one of=20 them are default. And as for self documentation? auto c3ptr =3D new C(default, /*timeout*/1); // use a shorter timeout=20 On Tuesday, 18 August 2015 15:03:38 UTC+1, Matthew Woehlke wrote: > > On 2015-08-18 06:22, David Rodr=C3=ADguez Ibeas wrote:=20 > > On Tue, Aug 18, 2015 at 10:42 AM, Arthur Tchaikovsky wrote:=20 > >> On Tuesday, 18 August 2015 10:34:39 UTC+1, Johannes Schaub wrote:=20 > >>> I would also wanna solve this with named arguments=20 > >>>=20 > >>> f(.b =3D 2, .d =3D 4)=20 > >>=20 > >> That genuinely seems pointless/silly to me, unless you can provide som= e=20 > >> convincing use cases.=20 > >=20 > > Any use case in which you may end up using 'default' in your proposal i= s=20 > a=20 > > use case for these two approaches. In the first case, it lets you focu= s=20 > on=20 > > what parameters are being configured (rather than defaulted) without=20 > having=20 > > to look at the function declaration, 'f(.b =3D 2, .d =3D 4)' is equival= ent=20 > to=20 > > 'f(default, 2, default, 4)' and I find the former easier to read.=20 > > It's also self-documenting. Consider Arthur O'Dwyer's examples; which is= =20 > easier to understand?=20 > > auto cptr =3D new C(); // usually the defaults are okay=20 > auto c2ptr =3D new C(2048); // use a bigger buffer=20 > auto c3ptr =3D new C(default, 1); // use a shorter timeout=20 > > - vs. -=20 > > auto cptr =3D new C{};=20 > auto c2ptr =3D new C{.buffer =3D 2048};=20 > auto c3ptr =3D new C{.timeout =3D 1};=20 > > There's a reason named arguments are so popular in Python :-).=20 > > > int a[10] =3D { [3] =3D 1, [2] =3D 5, 6 }; // what values are in the ar= ray?=20 > > This never happens because your example is ill-formed. Parameters that=20 > are neither named nor indexed must precede parameters that are. (In=20 > Python also, and for similar reasons.)=20 > > > Still, that might be easier than implementing named arguments. The=20 > "named"=20 > > arguments work nicely with a 'struct' since the type can be defined onl= y=20 > > once, but it leads to different sorts of confusion for functions that= =20 > may=20 > > have multiple declarations:=20 > >=20 > > void f(int a =3D 1, int b =3D 2);=20 > > void f(int b =3D 1, int a =3D 2); // redeclaration=20 > >=20 > > Which AFAIK is one of the issues that this form of proposals have faced= =20 > in=20 > > the past. I recall also some discussions in this forum (or maybe even= =20 > > papers?) aiming for a syntax that would allow this by enabling=20 > > warning/errors when a redeclaration used different names than the=20 > original=20 > > declaration but I did not see any of those prosper.=20 > > Probably there should be a way of naming parameters that "codifies" the= =20 > name, making it part of the ABI. And, yes, such redeclarations would be= =20 > disallowed.=20 > > --=20 > Matthew=20 > > --=20 ---=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. Visit this group at http://groups.google.com/a/isocpp.org/group/std-proposa= ls/. ------=_Part_439_656562220.1439915875035 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
I dear to disagree with the statement that the following e= xample is self documenting and easier to read:
auto c3ptr =3D new C{.tim= eout =3D 1};
How is this^^^ easier to read? And how is this self documenting? Only o= n surface it seems that it is, both, but when you really think about it it = is neither self documenting nor easy to read. Why? Because having only one = argument initialized in {}, that is .timeout, leaves us with question, is t= hat the only argument to that class, or are there any other? If so, are tho= se arguments inited by default? If so, how many? Etc.
Compared to this:<= br>auto c3ptr =3D new C(default, 1); // use a shorter timeout
It is immediately obvious how many arguments there are and which one of= them are default. And as for self documentation?
auto c3ptr =3D new C(d= efault, /*timeout*/1); // use a shorter timeout


On Tuesday, 18 August 2015 15:03:38 UTC+1, Matthew Woehlke wro= te:
On 2015-08-18 06:22, David = Rodr=C3=ADguez Ibeas wrote:
> On Tue, Aug 18, 2015 at 10:42 AM, Arthur Tchaikovsky wrote:
>> On Tuesday, 18 August 2015 10:34:39 UTC+1, Johannes Schaub wro= te:
>>> I would also wanna solve this with named arguments
>>>
>>> =C2=A0 f(.b =3D 2, .d =3D 4)
>>
>> That genuinely seems pointless/silly to me, unless you can pro= vide some
>> convincing use cases.
>
> Any use case in which you may end up using 'default' in yo= ur proposal is a
> use case for these two approaches. =C2=A0In the first case, it let= s you focus on
> what parameters are being configured (rather than defaulted) witho= ut having
> to look at the function declaration, 'f(.b =3D 2, .d =3D 4)= 9; is equivalent to
> 'f(default, 2, default, 4)' and I find the former easier t= o read.

It's also self-documenting. Consider Arthur O'Dwyer's examp= les; which is
easier to understand?

=C2=A0 auto cptr =3D new C(); =C2=A0 =C2=A0 =C2=A0 =C2=A0 // usually th= e defaults are okay
=C2=A0 auto c2ptr =3D new C(2048); =C2=A0 =C2=A0// use a bigger buffer
=C2=A0 auto c3ptr =3D new C(default, 1); // use a shorter timeout

- vs. -

=C2=A0 auto cptr =3D new C{};
=C2=A0 auto c2ptr =3D new C{.buffer =3D 2048};
=C2=A0 auto c3ptr =3D new C{.timeout =3D 1};

There's a reason named arguments are so popular in Python :-).

> int a[10] =3D { [3] =3D 1, [2] =3D 5, 6 }; // what values are in t= he array?

This never happens because your example is ill-formed. Parameters that
are neither named nor indexed must precede parameters that are. (In
Python also, and for similar reasons.)

> Still, that might be easier than implementing named arguments. =C2= =A0The "named"
> arguments work nicely with a 'struct' since the type can b= e defined only
> once, but it leads to different sorts of confusion for functions t= hat may
> have multiple declarations:
>=20
> void f(int a =3D 1, int b =3D 2);
> void f(int b =3D 1, int a =3D 2); =C2=A0// redeclaration
>=20
> Which AFAIK is one of the issues that this form of proposals have = faced in
> the past. =C2=A0I recall also some discussions in this forum (or m= aybe even
> papers?) aiming for a syntax that would allow this by enabling
> warning/errors when a redeclaration used different names than the = original
> declaration but I did not see any of those prosper.

Probably there should be a way of naming parameters that "codifies= " the
name, making it part of the ABI. And, yes, such redeclarations would be
disallowed.

--=20
Matthew

--

---
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.
Visit this group at http://groups.google.com/a/isocpp.org/group/std-proposals/<= /a>.
------=_Part_439_656562220.1439915875035-- ------=_Part_438_480074727.1439915875035-- .