Everything I agreed with and (hopefully) fixed has been removed without comment.

On Fri, Aug 14, 2020 at 9:02 AM <xxxxxx@ancell-ent.com> wrote:

This SRFI recommends that in multi-threaded programs the mask be set in the primordial thread before any other threads are created and never changed again.

The above and its repeat for set-current-directory! would be more clear with an Oxford comma at the end, "are created, and never changed again."  That helps to emphasize the very important last clause.

Oxford commas are of the form "A, B, and C".  They don't justify "A, and B", which is what you're asking for here.  Instead, I put the last three words in italics.
 
I also wasn't able to find anything in the POSIX reference saying CLOCK_REALTIME excludes leap seconds, while https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/time.h.html says WRT to the tm struct, "The range [0,60] for tm_sec allows for the occasional leap second."

Section A.4.16 <https://pubs.opengroup.org/onlinepubs/9699919799/xrat/V4_xbd_chap04.html#tag_21_04_16> makes it clear that "seconds since the Epoch" ignores leap seconds; it is not an actual count.

Leap seconds are ignored in order to guarantee that there are exactly 86400 seconds per day, a property that a great deal of software depends on.  You can represent a broken-out time with tm_sec == 60, but the Posix time functions will never hand you one.

 So for possible language, "... Universal Time), using CLOCK_REALTIME, which is defined in time.h."  And I suppose it might be profitably mirrored in the description of monotonic-time, despite it being obvious.

We don't talk about specific headers anywhere else.
 
The explanation at the end would be more clear if you add the typical examples, "is sometime turns either forward or backward, to for example implement daylight savings or summer time, or leap seconds."

None of those things affect the CLOCK_REALTIME clock.  Remember that it's ticking UTC-without-leap-seconds, not local civil time.  What does affect it is corrections for clock drift, either by NTP or by hand (or conceivably for testing purposes) and the text now points that out.

Everything is pushed.


John Cowan          http://vrici.lojban.org/~cowan        xxxxxx@ccil.org
Thor Heyerdahl recounts his attempt to prove Rudyard Kipling's theory
that the mongoose first came to India on a raft from Polynesia.
        --blurb for Rikki-Kon-Tiki-Tavi