Path: csiph.com!eternal-september.org!feeder.eternal-september.org!mx02.eternal-september.org!.POSTED!not-for-mail
From: Keith Thompson <kst-u@mib.org>
Newsgroups: comp.lang.c
Subject: Re: C pointer semantics with dynamically allocated struct
Date: Mon, 17 Aug 2015 16:42:20 -0700
Organization: None to speak of
Lines: 37
Message-ID: <lnwpwt4h43.fsf@kst-u.example.com>
References: <mqdg0t$rv2$1@speranza.aioe.org> <mqi018$eaf$1@dont-email.me> <003ef111-3124-49dc-b5cc-4d1bc0cc258c@googlegroups.com> <D6wzx.1885$K21.782@fx18.iad> <d57927ae-20ac-4ad4-b82b-b232c745e0c9@googlegroups.com> <Trxzx.34$KQ.31@fx03.iad> <e6608a24-899b-443f-8515-d19fd2c0030e@googlegroups.com> <mqmec8$5om$1@speranza.aioe.org> <mqqfjf$hnv$1@speranza.aioe.org> <mqqj0d$hec$1@dont-email.me> <mqreg9$gb2$1@speranza.aioe.org> <mqscrm$gce$1@dont-email.me> <mqsdmr$fo9$1@speranza.aioe.org> <mqsf72$od8$1@dont-email.me> <for-20150817133224@ram.dialup.fu-berlin.de> <mqsitu$riv$1@speranza.aioe.org> <79d68564-6e2e-447d-a73b-7b19d4e8bc66@googlegroups.com> <mqske6$v37$1@speranza.aioe.org> <ln4mjx7u3r.fsf@kst-u.example.com> <mqtfmb$av$1@speranza.aioe.org> <55D25210.20306@verizon.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Injection-Info: mx02.eternal-september.org; posting-host="945944de09706c9b4e29b53c9d2efdc2"; logging-data="6073"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX18aNBsDPTo3a8CUrvjtQwVE"
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/23.1 (gnu/linux)
Cancel-Lock: sha1:AY74rq8IFij/vLnCWlIxEySa/Ys= sha1:z4eSEC6p306yWqGdROQVrIyIkAg=
Xref: csiph.com comp.lang.c:67652

James Kuyper <jameskuyper@verizon.net> writes:
> On 08/17/2015 04:20 PM, Morten W. Petersen wrote:
>> On 17.08.2015 18:33, Keith Thompson wrote:
>>> "Morten W. Petersen" <morphex@gmail.com> writes:
>> [...]
>>>> Macros with parameters for example, why not have a compiler that
>>>> must support inline?  No more complicated macro implementations for
>>>> compiler writers, no more headaches for developers.
>>>
>>> Because there are plenty of macros that cannot be expressed as inline
>>> functions.  A simple example:
>>>
>>>      #define MAX(x, y) ((x) > (y) ? (x) : (y))
>>>
>>> An inline function would have to specify the type of the arguments.
>>> (Of course the macro has the drawback that one of the arguments
>>> will be evaluated twice.)
>> 
>> Well, it might not be possible to express a function in
>> exactly the way as a macro, but something equivalent
>> should be possible.
>
> Have you actually thought that through? The key difference between a
> macro and a function is that the function necessarily has a single
> specific type for each parameter, and for it's return type.
[...]

Mostly true.  Variadic functions (like printf) can take arguments
of varying types.  But even so, the return type is fixed, so you
can't write a max() function that's anywhere near as convenient as
the macro.

-- 
Keith Thompson (The_Other_Keith) kst-u@mib.org  <http://www.ghoti.net/~kst>
Working, but not speaking, for JetHead Development, Inc.
"We must do something.  This is something.  Therefore, we must do this."
    -- Antony Jay and Jonathan Lynn, "Yes Minister"