Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 22 Feb 2019 22:03:07 +0000
From:      bugzilla-noreply@freebsd.org
To:        bugs@FreeBSD.org
Subject:   [Bug 235946] column(1): Fill order (with/without -x) is incorrectly documented
Message-ID:  <bug-235946-227@https.bugs.freebsd.org/bugzilla/>

next in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D235946

            Bug ID: 235946
           Summary: column(1): Fill order (with/without -x) is incorrectly
                    documented
           Product: Documentation
           Version: Latest
          Hardware: Any
                OS: Any
            Status: New
          Severity: Affects Only Me
          Priority: ---
         Component: Manual Pages
          Assignee: bugs@FreeBSD.org
          Reporter: dana@dana.is
                CC: doc@FreeBSD.org

Created attachment 202276
  --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=3D202276&action=
=3Dedit
column(1) documentation fix

The documentation for column(1) says that 'Rows are filled before columns' =
and
that the -x option 'Fill[s] columns before filling rows'. The language dates
back to at least 4.4BSD, and it matches the names of the functions responsi=
ble
for the corresponding output modes in the source code. However, i don't thi=
nk
that most people would interpret that wording the way the original develope=
rs
apparently did:

% printf '%s\n' aaa bbb ccc ddd eee | column -c24
aaa     ccc     eee
bbb     ddd
% printf '%s\n' aaa bbb ccc ddd eee | column -xc24
aaa     bbb     ccc
ddd     eee

To me, the default behaviour is clearly filling the columns first, whilst t=
he
-x behaviour is filling the rows first.

The developers of the util-linux[1] and OpenBSD[2] implementations of colum=
n(1)
evidently agree, since their man pages now state that the default behaviour=
 is
to fill columns first.

I've attached a patch that changes the man page to bring it in line with
util-linux and OpenBSD.

When util-linux fixed their documentation, they also updated the source cod=
e to
match. OpenBSD did not, maybe because simply swapping the function names wo=
uld
make it very confusing to compare the source to other/historical BSDs. I've
gone the OpenBSD route here, but i did add a comment about it.

PS: Sorry if i've misunderstood anything; this is my first time submitting a
bug/patch to FreeBSD directly.

[1]
https://github.com/karelzak/util-linux/blob/2698f9ba887cb7e4a204ada72016b1c=
1192b17c1/text-utils/column.1

[2]
https://github.com/openbsd/src/blob/c37de4bcf15ea51b5fae88c6b1911bf61532958=
6/usr.bin/column/column.1

--=20
You are receiving this mail because:
You are the assignee for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-235946-227>