docs/100494

[Maxim Konovalov]

The following reply was made to PR docs/100494; it has been noted by GNATS.

From: Maxim Konovalov <maxim at macomnet.ru>
To: bug-followup at freebsd.org
Cc:  
Subject: docs/100494
Date: Thu, 28 Sep 2006 10:35:06 +0400 (MSD)

 Added to the audit trail.
 
 -- 
 Maxim Konovalov
 
 ---------- Forwarded message ----------
 Date: Tue, 26 Sep 2006 03:09:49 -0400 (EDT)
 From: waldeck at gmx.de
 To: undisclosed-recipients:  ;
 Subject: Re: Re: Re: docs/100494
 
 Hi Maxim,
 
 you did not answer my last email!
 
 > I'm not sure I understand your quiestion completely.  In UFS2 there is
 > a lazy inode allocation mechanism but it is irrelevant to the topic
 > because st_birthtime is an attribute of the file the inode refers.
 
 Summary:
 1. birthtime is NOT a standard Un!x feature.
 2. The biological term "birth" is misleading and unsuitable!
 3. A clear and evident definition is required!
 4. During the file creation a inode is allocated.
 5. This time is saved in st_birthtime.
 
 man 1 stat should be:
 > a, m, c, B
 >		     The time file was last accessed or modified, of when the
 > 		     inode was last changed, or the time of file creation
 >                    (the so called birth time of the inode).
 
 instead of:
 > 		     inode was last changed, or the birth time of the inode.
 
 man 2 stat should be:
 > 		     st_birthtime  Time when the file was created and the inode was allocated.
                                                  ^^^^
 instead of:
 > 		     st_birthtime  Time when the inode was created.
 
 Thanks!
 
 Markus