Path: csiph.com!eternal-september.org!feeder.eternal-september.org!reader01.eternal-september.org!.POSTED!not-for-mail
From: Keith Thompson <Keith.S.Thompson+u@gmail.com>
Newsgroups: comp.lang.c
Subject: Re: Ways to avoid global variables
Date: Sun, 05 Apr 2020 16:12:50 -0700
Organization: None to speak of
Lines: 76
Message-ID: <875zedplvx.fsf@nosuchdomain.example.com>
References: <r6c5ie$5b9$1@dont-email.me> <r6cr3a$bt1$1@dont-email.me> <global-20200405160642@ram.dialup.fu-berlin.de> <r6cua6$1o0$1@dont-email.me> <variables-20200405170535@ram.dialup.fu-berlin.de>
Mime-Version: 1.0
Content-Type: text/plain
Injection-Info: reader02.eternal-september.org; posting-host="9aca32966cb66c1a01b4dff59d5f3d44"; logging-data="19654"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX18qni7PaivmRs3ginXjwSdy"
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.3 (gnu/linux)
Cancel-Lock: sha1:y4HHRztppI6LCjEaT3NPY0MlFYk= sha1:XKBM8eTxOnqe7RnrG7Ekj9n//JE=
Xref: csiph.com comp.lang.c:151106

ram@zedat.fu-berlin.de (Stefan Ram) writes:
> James Kuyper <jameskuyper@alumni.caltech.edu> writes:
>>On 4/5/20 11:07 AM, Stefan Ram wrote:
>>>Pointers can be the null pointer or point to entities
>>>(functions or objects), but not to variables.
>>I'm curious - what do you consider the definition of "variable" to be,
>>that justifies your claim that pointers can't point at them?
>
>   A variable contains all the information that is given by the
>   declaration of a name and by the scope of this declaration.
>   This encompasses the name itself, its type, and also, for
>   example, whether this is an automatic variable.
>
>   For example, in the block,
>
> { int x; }
>
>   , a variable with the name "x" and the type "int" and
>   automatic storage duration is declared.

Like most definitions of the word "variable" I've seen, this makes
statements about variables or about what they do, but it doesn't
tell me what a variable *is*.

I think most programmers would assume that a "variable" is a kind
of object.  Others might say that it's some kind of abstract thing
that can refer to an object.  My best guess from your wording is
that a "variable" is a compile-time symbol table entry.

Given:

    int x;
    int *p = &x;

I would say that p points to a variable -- or, since C doesn't say
what a "variable" is, that p points to a named object.  It's certainly
true that p does not give access to the name "x".

>   Akin to:
>
> |In C, all variables must be declared before they are used,
> |usually at the beginning of the function before any
> |executable statements. A declaration announces the
> |properties of variables; it consists of a name and a list of
> |variables, such as
> |int fahr, celsius;
> |int lower, upper, step;
> The C Programming Language, Second Edition
>
>   When one has a pointer to an object that is bound to a name,
>   one cannot reach information about the name.
>
>   When a pointer would be able to point to a variable, I would
>   then also expect to be able to obtain the name via the pointer.
>
>   Via recursion, several objects can be bound to a variable that
>   is declared once. This shows that there is not a 1:1 correspondence
>   between variables and objects.
>
>>In my experience, by far the most common problem with global variables
>>IS their excessive visibility, so I would not describe that as "merely".
>
>   Global variables /hide/ dependencies of functions.
>
>   For example, the behavior of
>
> void print( int const n ){ printf( "%d\n", n ); }
>
>   depends on "stdout", but this is not shown in the interface.
>
>

-- 
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
Working, but not speaking, for Philips Healthcare
void Void(void) { Void(); } /* The recursive call of the void */