Re: origin of "void *" (was: [PATCH v1 19/25] man/man3type/void.3type: HISTORY: Update first POSIX appearance of void(3type)

[Alejandro Colomar <alx@kernel.org>]

[-- Attachment #1: Type: text/plain, Size: 3136 bytes --]

On Sat, Jan 10, 2026 at 01:07:03PM +0100, Alejandro Colomar wrote:
> Hi Branden,
> 
> On Sat, Jan 10, 2026 at 05:52:28AM -0600, G. Branden Robinson wrote:
> > Hi Alex,
> > 
> > At 2026-01-10T12:30:13+0100, Alejandro Colomar wrote:
> > [rearranged]
> > > On Sat, Jan 10, 2026 at 07:57:49AM +0000, Seth McDonald wrote:
> > > > I know that the void pointer type was not mentioned or used in
> > > > POSIX.1-1988.  Instead, the standard used the char pointer type for
> > > > pointers to generic data.  POSIX.1-1990 seems to implicitly require
> > > > void pointers - at least for conformance to "POSIX.1, C Language
> > > > Binding (C Standard Language-Dependent System Support)" - by
> > > > including them in some function prototypes.
> > > 
> > > > On Friday, 9 January 2026 at 20:34, Alejandro Colomar
> > > > <alx@kernel.org> wrote:
> > > Hmmm, since POSIX.1-1988 was already written after some C89 draft, it
> > > seems void* was incorporated to C89 in a late draft.
> > [...]
> > > Yup, K&R C 1st ed. is too old and I wouldn't expect it to have
> > > void*.  I have a physical copy of K&R C 2nd ed. (1988) and it does
> > > have void*.
> > [...]
> > > Being a scan [of 1e], it could be authentic.  I don't have a physical
> > > copy of the first edition, so can't compare.
> > [...]
> > > If anyone else reading the list knows anything, it would be welcome.
> > > 
> > > Be careful of online copies of K&R C, though.  I once found one of the
> > > second edition, and it was fake (by comparing it to the actual copy).
> > > 
> > > Much better to get the physical copy.  Although the first edition seems
> > > difficult to find.
> > 
> > I have a physical copy of K&R 1e, but unfortunately it is in storage and
> > not easy to access.  However...
> > 
> > I have before me as I type this a copy of Tom Plum's _Notes on the Draft
> > C Standard_, copyright 1987 (Plum Hall, ISBN 609-927-3770).  Publishers
> > treat years in copyright notices the way automakers treat them in
> > vehicle models, so I'd guess this puts the date of publication anywhere
> > from 4Q1986 to 3Q1987.
> > 
> > In Chapter 2, "Comparison with Earlier C", on p. 24, it says:
> > 
> > 	"A pointer to any object type may be converted to a ‘‘generic’’
> > 	pointer and back again without change.  The syntax for generic
> > 	pointer is void *."
> > 
> > Tom Plum was a noteworthy figure in C standardization circles; his name
> > shows up in much literature on the subject.  I'd regard him as an
> > authority.
> 
> Thanks!
> 
> I've now checked the rationale for C89, and it says that
> 
> 	Several new types have been added:
> 
> And mentions 'void', 'void*', and 'signed char'.  It seems it was an
> invention of C89.

Forgot to add link:
<https://www.open-std.org/jtc1/sc22/wg14/www/C89Rationale.pdf>

> 
> 
> Hve a lovely day!
> Alex
> 
> > 
> > https://plumhall.com/newsite/phall.html
> > 
> > Looks like he passed away last year.  :(
> > 
> > Regards,
> > Branden
> 
> 
> 
> -- 
> <https://www.alejandro-colomar.es>



-- 
<https://www.alejandro-colomar.es>

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]