Bruce Evans <b...@optusnet.com.au> wrote in <20130912222312.k1...@besplex.bde.org>:
br> > On 9/12/2013 6:36 AM, Hiroki Sato wrote: br> >> I think this kind of changes need a consensus because several POSIX br> >> functions use "filedes" in the specification document. r254484 by br> >> pjd was a similar change (s/type/af/ in gethostbyaddr()). br> >> br> >> In SUSv4, fdopen() uses "filedes" and openat() uses "fd", for br> >> example. Consistency throughout our manual pages is generally good. br> >> However, I also see the benefit of using the same expression as the br> >> specification even if it is inconsistent. What do you think? br> br> Does it really use "filedes"? POSIX still never uses this in the 2007 br> draft (austin-d2r.pdf). It uses "fildes" for fdopen(), but "fd" for br> fdopendir() and openat(). It still uses "fd" for posix_fadvise() and br> posix_fallocate(). I now think that the "fd"s in POSIX are just br> style bugs. The normal "fildes" had only rotted to "fd" in 2 places br> in 2001, but rotted much further in 2007. No, it was fildes, not filedes, as you pointed out. It seems that my fingers preferred "e" over "d" after "fil"... -- Hiroki
pgplyT9qHzBnG.pgp
Description: PGP signature
