This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFC] Canonical Standards Names


On 06/21/2017 09:45 AM, Joseph Myers wrote:
> there are three things we need to decide: (a) 
> what standards need names for @standards at all (e.g., do BSD and SVID 
> need separate names); (b) given which standards need names, what those 
> names should be;

... [(c) not addressed below]

> (a) is important before 
> we can know what an answer for (b) should look like.

Here is what currently exists:

$ grep -h '^@standards' manual/*.texi \
    | sed -r 's/^@standardsx?\{(([^,]+), ){1,2}.*/\2/' | sort -u
???
BSD
C90
C99
GNU
IPv6 basic API
IPv6 Basic API
ISO
ISO/Amend1
ISO/IEC TS 18661-1
ISO/IEC TS 18661-3
LFS
Linux
Linux???
Obsolete
POSIX
POSIX.1
POSIX.1b
POSIX.1c
POSIX.1g
POSIX.2
SunOS
SUNRPC
SVID
TS 18661-1:2014
Unix
Unix98
Unknown origin
X/Open
XOPEN
XPG
XPG2
XPG4
XPG4.2

AIUI, @standards is meant to imply correlation with feature test macros
as a service to programmers (as opposed to serving the interests of
historians), so DEFAULT is my preferred choice, since __USE_MISC =
_DEFAULT_SOURCE (= _(BSD|SVID)_SOURCE), and it's most to the point.

I think the use of generic standards to imply a forebear should be
avoided; e.g., ISO = C90, or POSIX = POSIX.1-2001.  XPG may pose an
issue there, however.

Rical


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]