Update to match.... We're slow but we get there eventually!
NFC for any of these programs, struct timeval and struct timespec
are the same size, and only the tv_sec field of boottime is used,
and that's unchanged.
Use at most half of ttywidth for WHERE field (hostname) instead
of a hardcoded maximum width. (XXX: this is still not a good
default for displays under 80 columns. Should probably count
the width of all the other columns, using 8 for the WHAT field.)
that things don't wrap. Add -w flag to enable full output. Delete mention
of the old and not supported flags. This was 11 years ago, and very few
people even remember what these flags did.
indicated by the pid in struct utmpx [or ftpd hack], and the one that is
associated with the controlling tty, is in the foreground process group,
and has most cpu. If the second one exists, use that one, else use the first
one.
not need it since it can get the pid from the utmpx structure.
- If the line is not found, and the pid is present (from utmpx) use that to
determine the process information for that particular entry.
Get things right for tty names of length UT_LINESIZE. (FreeBSD PR 3638)
If the utmp login time is greater than the tty atime, use it to calculate
the idle time instead of the atime.
Fix -n with 16-chars hostnames (FreeBSD PR 11286).
Domain names are case insensitive.
__CONCAT("foo","bar");
actually works to concantate strings, it's because the preprocessor expands
it into "foo""bar" as separate strings, and then ANSI string concatenation
is performed on that. It's more straightforward to just use ANSI string
concatenation directly, and newer GCCs complain (rightly) about misuse
of token pasting.
process accordingly, so that w parses entries made by "ftpd -U" correctly
the new code is conditionalized upon define SUPPORT_FTPD_UTMP, turned
on by default in w(1)'s Makefile
This addresses bin/11095 by Daniel Hagerty.
