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: "Why the C Language Will Never Stop You from Making Mistakes" by JeanHeyd Meneide
Date: Sun, 16 Aug 2020 10:50:17 -0700
Organization: None to speak of
Lines: 34
Message-ID: <87wo1yeazq.fsf@nosuchdomain.example.com>
References: <rh1du1$q82$1@dont-email.me> <877du33by4.fsf@nosuchdomain.example.com> <alpine.BSF.2.22.395.2008121503080.21003@slashem.me> <rh2475$r12$1@reader1.panix.com> <87r1sb1fu5.fsf@nosuchdomain.example.com> <rh2lpf$5v3$1@reader1.panix.com> <87k0y312ua.fsf@nosuchdomain.example.com> <kD9ZG.114542$mK4.89063@fx03.iad> <rh3feq$n8h$1@dont-email.me> <SwxZG.133451$Vp4.64818@fx44.iad> <rh6eo9$2id$1@dont-email.me> <GRCZG.121967$TO4.71714@fx05.iad> <rh77um$5qk$1@dont-email.me> <ec3512a0-fabf-4291-85de-611ca70104b0o@googlegroups.com> <88804d79-6f6f-427a-ae42-33fb0d07dd9ao@googlegroups.com> <Bta_G.50563$5l1.20045@fx10.iad>
Mime-Version: 1.0
Content-Type: text/plain
Injection-Info: reader02.eternal-september.org; posting-host="0c763dcea4bade25e66f8f642b4cd05d"; logging-data="19543"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/2KJ/mTJQsGiHsZ8Um83sa"
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.3 (gnu/linux)
Cancel-Lock: sha1:43HxaPzKwUn9jaG/JSOx1F710LQ= sha1:BYgr26DwJCdHH98GfuOfFt0e4qk=
Xref: csiph.com comp.lang.c:153673

Richard Damon <Richard@Damon-Family.org> writes:
[...]
> Again, the want to remove the cast comes from C90 mindsets, and poorly
> segmented code. Starting with C99, best practices was to try to
> initialize things as soon as they were created, and initialization is
> the most type sensitive part of a program, so it is best if the type was
> mentioned in that section of code.

Yes, one of the reasons to prefer the castless version became (somewhat)
obsolete with C99's removal of implicit int.

> Again, what HARM does the cast impose (particularly if hidden behind a
> well named macro that avoids the repetition.)

My main objection to using a cast is conceptual.  A pointer cast
in C is too unrestrained.  I can be used to (try to) reinterpret
memory of some time as if it were of a completely different type,
perhaps with arbitrary bad results.  The fact that the same syntax
is used for a harmless and save conversion means that when I see a
cast, particularly a pointer cast, I have to convince myself that
it's safe.

And in this case it's simply unnecessary.

    int *p = (void*)malloc(sizeof (int));
    uint32_t n = (uint32_t)42;

Both casts are equally "safe" -- and equally unnecessary.  And I find
the code simpler and easier to read without them.

-- 
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 */