Re: "A diagram of C23 basic types"

Keith Thompson <Keith.S.Thompson+u@gmail.com> wrote:

antispam@fricas.org (Waldek Hebisch) writes:

Tim Rentsch <tr.17687@z991.linuxsc.com> wrote:

[...]

Not always practical.  A good example is the type size_t.  If a
function takes an argument of type size_t, then the symbol size_t
should be defined, no matter which header the function is being
declared in.

>
Why?  One can use a type without a name for such type.

 
Convenience and existing practice.  Sure, an implementation of
<string.h> could provide a declaration of memcpy() without making
size_t visible, but what would be the point?

Cleanliness of definitions?  Consistency?  Fragment that you
replaced by [...] contained a proposal:

Every identifier should be declared in exactly one home header,
and no other header should provide that definition.

That would be pretty clean and consistent rule: if you need some
standard symbol, then you should include corresponding header.

Tim claimed that this in not practical.  Clearly C standard changed
previous practice about headers, so existing practice is _not_
a practical problem with adapting such proposal.  With current
standard and practice one frequently needs symbols from several
headers, so "convenience" is also not a practival problem with
such proposal.  People not interested in clean name space can
define private "all.h" which includes all standard C headers
and possibly other things that they need, so for them overhead
is close to zero.

--
                              Waldek Hebisch