Package: util-linux
Version: 2.13.1.1-1
Severity: normal

Hello LaMont,

setarch man page is deficient in many ways:

1) the SYNOPSYS is broken:
SYNOPSIS
       setarch <arch> [options] [program [arguments]]

       arch [options] [program [arguments]]

why is there an "arch" line ?

2) it does not include reference to linux32/linux64 while these manpages are
linked to it.
3) it states that "This utility currently only affects the output of uname -m."
and then goes on to state the opposite.
4) it lists a lot of options (SHORT_INODE, ADDR_LIMIT_32BIT, without
explaining what they do or where they are documented.

Cheers,

* Bill Allombert <ballombe@debian.org> [200503 22:31]:

"arch" refers to a number of aliases, incl. linux32/linux64 as you
mentioned below. But also i686, s390, ppc64le and what have you.
This generic "arch" trick probably avoids listing all variants
and/or having arch-specific manual pages.

See above.

TTBOMK - by default - it only changes the uname -m bit. The options
turn on different changes.

These appear to have been given short explanations in the meantime.

If there are further suggestions, please send patches to upstream -
they are quite receptive.

Thanks,
Chris

reopen 530011
quit

Hello Chris,

You forget to CC me while closing this bug.
Bug submitters must always be CCed since they do not receive
mail to xxx@bugs.debian.org, only the maintainer does.

Thanks for the explanation!

This should be explained and clarified in the manpage.
arch is typeset in the same way as setarch, so the reader
expects this to refer to  /usr/bin/arch, but this is not the case.
/usr/bin/arch is provided by coreutils and is unrelated to setarch,
and has different options.
Instead it should refer to linux32/linux64 and state other
aliases might exist.

Cheers,

* Bill Allombert <ballombe@debian.org> [200504 12:42]:

While your second statement is correct, I did not forget to CC you,
because I mailed xxx-done@. https://www.debian.org/Bugs/Developer#closing
clearly says that messages to xxx-done@ also get sent to the
submitter, therefore you.

Probably. Do you care enough to send a patch to upstream?
That would greatly help with getting this done.

Thanks,
Chris