On Jan 13 17:22, Jon Turney wrote: > On 13/01/2025 13:38, Corinna Vinschen wrote: > > On Jan 10 17:01, Brian Inglis wrote: > > > Move circular F/Ff/Fl and similar functions to put > > > base entries and -f -l variants on the same line. > > > > > > Signed-off-by: Brian Inglis <brian.ing...@systematicsw.ab.ca> > > > --- > > > winsup/doc/posix.xml | 336 +++++++++++-------------------------------- > > > 1 file changed, 84 insertions(+), 252 deletions(-) > > > > Hmm. This makes more sense than the previous patch. > > > > However, it doesn't make sense if only the math functions are affected > > If you want to do this systematically, you'd have to group them > > following the Open Group Base Spec Issue 8. > > > > That would also imply merging stuff like > > > > iswalnum/iswalnum_l -- page 1280 > > le16toh/le32toh/le64toh -- page 1327 > > localtime/localtime_r -- page 1354 > > mlock/munlock -- pages 1433, 1488 > > sig2str/str2sig -- page 2040 > > > > etc. etc. > > > > Nevertheless, while this has a certain charm and I don't see > > a disadvantage, I'd like to get Jon's input to this as well. > > Well, I'd just go for one function per line, in alphabetical order, because > it's unambiguous if we've done that correctly. > > Otherwise, it devolves into arguments about "I think it looks nicer this > way", which... well, "de gustibus non est disputandum".
Good point. Yeah, let's keep them one per line in ASCII order. Corinna
