cvs commit: src/sys/ufs/ufs ufs_vnops.c

[Bruce Evans]

On Thu, 1 Jun 2006, John Baldwin wrote:

> On Thursday 01 June 2006 06:01, Alexey Dokuchaev wrote:
>> On Thu, Jun 01, 2006 at 10:49:50AM +0100, Ceri Davies wrote:
>>> @@ -69,6 +69,10 @@
>>>  the file must be open for writing.
>>>  .Sh RETURN VALUES
>>>  .Rv -std
>>> +If the file to be modified is not a directory or
>>> +a regular file, the
>>> +.Fn truncate
>>> +call will return the value 0.
>>
>> Doesn't "value of 0" sound better?
>
> Not to me, though I can't explain why.  I think the phrase "X will return the
> value Y" is common in man pages though.

"will return" sounds strange to be.

Normal is "Upon successful completion, the value 0 is returned...".
This is part of what ".Rv -std" expands to.

POSIX says "Upon successful completion, ftruncate( ) shall return 0...".

The POSIX wording is better.  "the value 0" says nothing more than "0",
and "returns" is clearer than "is returned".  Saying "the value 0" is
apparently a hack to give the clause a subject (or is it an object? --
I think the value is the object convoluted to a subject or vice versa).

FreeBSD has a deshallify.sh script to adjust the POSIX wording.  It
does s/shall return/returns/.

Bruce