220 23733 article Path: news.gmane.org!not-for-mail From: Patrice Roy Newsgroups: gmane.comp.lang.c++.isocpp.proposals Subject: Re: Re: Default tuple-like access Date: Thu, 14 Jan 2016 21:02:04 -0500 Lines: 195 Approved: news@gmane.org Message-ID: References: <569583AD.9050208@wanadoo.fr> <5697E783.2020509@wanadoo.fr> <5698067B.3010706@wanadoo.fr> Reply-To: std-proposals@isocpp.org NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: multipart/alternative; boundary=001a113f9b442d4099052955ca8e X-Trace: ger.gmane.org 1452823332 10980 80.91.229.3 (15 Jan 2016 02:02:12 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Fri, 15 Jan 2016 02:02:12 +0000 (UTC) To: std-proposals@isocpp.org Original-X-From: std-proposals+bncBDSZXMHWZIKBBHFG4G2AKGQEKQJ66HA@isocpp.org Fri Jan 15 03:02:08 2016 Return-path: Envelope-to: gclcip-std-proposals@m.gmane.org Original-Received: from mail-oi0-f72.google.com ([209.85.218.72]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1aJtiJ-0005U2-7M for gclcip-std-proposals@m.gmane.org; Fri, 15 Jan 2016 03:02:07 +0100 Original-Received: by mail-oi0-f72.google.com with SMTP id x185sf1034338330oia.3 for ; Thu, 14 Jan 2016 18:02:06 -0800 (PST) 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:date:message-id:subject:from:to :content-type: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=CW+fRO0/CItFs4oSZJnHQ3UHv7JkZGAE87wtjqTVpBc=; b=yGwNlvXK8dMmj8lGWXx1b+g7BEcMr028D1ezC0G+ttCVz0a7MwWI9LbjNNFPLUpdn/ k7ccnH3JjYcgVMN3r5u+t82w67I/zNjE56dDq09XfxoCYRUE6vQ+K7N4yjQygAgA2Lvf jAnKqTz3gW2rpEKcA++U4GfEapL0dkfoxgiT+xlYzYxMtE7XwOCSAjnkWDMIEfBkt8gU 1yI7VAIMBWwwLiZZt6O3LFvZxa9KewCd7ODn5s1KL9bPOV3g8ieDlvlKkVnkcFBFYAb6 HM0hSWmC3D4S+1yT+XOV0Jh8jIIkpFvXY7DbqGDctNHNMYsr/50LjMJoA2OLyGJ3eOir fYQQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type: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=CW+fRO0/CItFs4oSZJnHQ3UHv7JkZGAE87wtjqTVpBc=; b=d8uyRZ4DRm8vC7lj67eFXm8m6FuDBhCt4hFWJ3pVV+F50BqTJX/8Bugd80VgfDINEk IobyqhJUL3UtnzT5eDEbI7qlhuqDHWNe18rJ552KxQB1UlZJbpPqr9br27TXcdmgPNTC SGq02/xcWtJ6OoL3qWEIVZ8W72UMfy95Lpjd7LjqjcPT/OfiH/t/lEW+nIabNOrsGcE/ gvQuaNkxkMqXEEOgfSP/LsnJtzB8tf/L5EBmg24lRP5hpOTk+Sg4Hwr92plr077rslTc VS7hhbJJG7rykQ4+JBAd9ewh7WZ1FA6bKv/AE8VkBPiviYCGBXC8Z9hh1/omkt1ULnfQ WBHQ== X-Gm-Message-State: ALoCoQm+2Pi07kYaOk+QQyR2a+w1Ikb6Yd6q8I7he22KH/L7i4mXoepJRXpL1oCI/1Q7W7wM7FvGTAqvNRW366E/yK/qEUXXqA== X-Received: by 10.107.152.10 with SMTP id a10mr7473728ioe.28.1452823326028; Thu, 14 Jan 2016 18:02:06 -0800 (PST) X-BeenThere: std-proposals@isocpp.org Original-Received: by 10.107.129.163 with SMTP id l35ls688551ioi.95.gmail; Thu, 14 Jan 2016 18:02:04 -0800 (PST) X-Received: by 10.107.2.65 with SMTP id 62mr7117139ioc.78.1452823324705; Thu, 14 Jan 2016 18:02:04 -0800 (PST) Original-Received: from mail-io0-x232.google.com (mail-io0-x232.google.com. [2607:f8b0:4001:c06::232]) by mx.google.com with ESMTPS id v15si887884igd.22.2016.01.14.18.02.04 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Thu, 14 Jan 2016 18:02:04 -0800 (PST) Received-SPF: pass (google.com: domain of patricer@gmail.com designates 2607:f8b0:4001:c06::232 as permitted sender) client-ip=2607:f8b0:4001:c06::232; Original-Received: by mail-io0-x232.google.com with SMTP id q21so472983188iod.0 for ; Thu, 14 Jan 2016 18:02:04 -0800 (PST) X-Received: by 10.107.10.88 with SMTP id u85mr7955270ioi.5.1452823324473; Thu, 14 Jan 2016 18:02:04 -0800 (PST) Original-Received: by 10.79.6.87 with HTTP; Thu, 14 Jan 2016 18:02:04 -0800 (PST) In-Reply-To: X-Original-Sender: patricer@gmail.com X-Original-Authentication-Results: mx.google.com; spf=pass (google.com: domain of patricer@gmail.com designates 2607:f8b0:4001:c06::232 as permitted sender) smtp.mailfrom=patricer@gmail.com; dkim=pass header.i=@gmail.com; dmarc=pass (p=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-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:23733 Archived-At: --001a113f9b442d4099052955ca8e Content-Type: text/plain; charset=UTF-8 The risks mostly come from program-generated tuples / variadics; these are the ones that will get long. I'm also uncomfortable with beliefs of low number of types / elements... Things such as tuples resulting from merging tuples, for example. Let's play it safe and plan for survival in the face of big numbers. 2016-01-14 17:21 GMT-05:00 Nevin Liber : > On 14 January 2016 at 16:11, Matthew Woehlke > wrote: > >> On 2016-01-14 16:40, Nevin Liber wrote: >> > On 14 January 2016 at 15:04, Matthew Woehlke wrote: >> >>> I don't see how implement the tuple_size you are suggesting >> efficiently. >> >>> How will you find the N? trying with get<0>, get<1>, get<2>, ... >> get, >> >>> up to a failing get? >> >> >> >> Yes, exactly. I don't see a serious efficiency problem here; I expect >> >> that in most cases, the number will be small, >> > >> > Please don't make that assumption. It reminds me of life before >> variadic >> > templates. >> >> I'm really struggling to think of why you would have a (non-array) >> tuple-like that has more than a few dozen members. > > > A tuple-like data structure to capture an expression template, for > instance. It can be arbitrarily large. > > Don't engineer in limits if you don't have to, and don't engineer in O(N) > algorithms (even at compile time) if you don't have to. There is no reason > not to expose it. What is the actual objection to doing so, other than we > aren't providing the absolute barest possible minimal set of operations by > doing so. > > >> That's a *very* >> complicated complex type. (Array-like types don't count; you'll almost >> certainly specialize tuple_size for those.) >> >> > One use case for knowing N is when you want to get things in reverse >> > order. Users shouldn't have to go through hoops just to figure out how >> to >> > do that, nor write code that requires O(N) template instantiations to >> > calculate it, because when they get it wrong, they'll still get horrible >> > error messages. >> >> ...but it *doesn't* "require O(N) template instantiations". That's just >> the case *if you don't provide a specialization*. If you have a type for >> which N is large, then provide your own tuple_size. > > > Why should I have to? This is why people rightly complain about C++. We > really don't have to make everything hard to accomplish. > -- > Nevin ":-)" Liber > +1-847-691-1404 > > -- > > --- > 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 > https://groups.google.com/a/isocpp.org/group/std-proposals/. > -- --- 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 https://groups.google.com/a/isocpp.org/group/std-proposals/. --001a113f9b442d4099052955ca8e Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
The risks mostly come from program-generated tuples / vari= adics; these are the ones that will get long. I'm also uncomfortable wi= th beliefs of low number of types / elements... Things such as tuples resul= ting from merging tuples, for example. Let's play it safe and plan for = survival in the face of big numbers.
2016-01-14 17:21 GMT-05:00 Nevin Liber <= nevin@cplusplusguy.com>:
On 14 January 2016 at 16:11, Matthew Woehlk= e <mwoehlke.floss@gmail.com> wrote:
On 2016-01-14 16:40, Nevin Liber wrote:
> On 14 January 2016 at 15:04, Matthew Woehlke wrote:
>>> I don't see how implement the tuple_size you are suggestin= g efficiently.
>>> How will you find the N? trying with get<0>, get<1>= ;, get<2>, ... get<N>,
>>> up to a failing get<N+1>?
>>
>> Yes, exactly. I don't see a serious efficiency problem here; I= expect
>> that in most cases, the number will be small,
>
> Please don't make that assumption.=C2=A0 It reminds me of life bef= ore variadic
> templates.

I'm really struggling to think of why you would have a (non-arra= y)
tuple-like that has more than a few dozen members.

A tuple-like data structure to capture an expression templat= e, for instance.=C2=A0 It can be arbitrarily large.

Don't engineer in limits if you don't have to, and don't engi= neer in O(N) algorithms (even at compile time) if you don't have to.=C2= =A0 There is no reason not to expose it.=C2=A0 What is the actual objection= to doing so, other than we aren't providing the absolute barest possib= le minimal set of operations by doing so.
=C2=A0=
That's a *very*
complicated complex type. (Array-like types don't count; you'll alm= ost
certainly specialize tuple_size for those.)

> One use case for knowing N is when you want to get things in reverse > order.=C2=A0 Users shouldn't have to go through hoops just to figu= re out how to
> do that, nor write code that requires O(N) template instantiations to<= br> > calculate it, because when they get it wrong, they'll still get ho= rrible
> error messages.

...but it *doesn't* "require O(N) template instantiations&q= uot;. That's just
the case *if you don't provide a specialization*. If you have a type fo= r
which N is large, then provide your own tuple_size.

<= /div>
Why should I have to?=C2=A0 This is why people rightly com= plain about C++.=C2=A0 We really don't have to make everything hard to = accomplish.
--
<= div dir=3D"ltr">=C2=A0Nevin ":-)" Liber=C2=A0 <mailto:nevin@cplusplusguy.com= >=C2=A0 +1-847-691-1404

--

---
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 https://groups.google.com/a/isocpp.org/g= roup/std-proposals/.

--

---
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 https://groups.google.com/a/isocpp.org/group/std-proposals= /.
--001a113f9b442d4099052955ca8e-- .