bug#5941: tail bug

2010-04-13 Thread Eric Blake 
On 04/13/2010 02:16 PM, Eric Kever wrote:
 I've created a file 'foo', and used tail -f to follow the changes to
 that file.
 I then wrote 'test' to the file and saved it, and tail reported 'test',
 which is fine.
 I then deleted 'test' and saved the file, and tail reported 'tail: foo:
 file truncated', which is fine.
 I then wrote 'test' again and saved the file, and tail reported 'est'
 instead of 'test'.

Which version of coreutils?  Which platform?  The latest coreutils is
8.4, and there have been some fixes for bugs in tail within the last
year.  We need more information before we can either repeat this or
definitively state that you have encountered one of those already-fixed
bugs.

-- 
Eric Blake   ebl...@redhat.com+1-801-349-2682
Libvirt virtualization library http://libvirt.org



signature.asc
Description: OpenPGP digital signature

bug#5941: tail bug

2010-04-13 Thread Andreas Schwab 
Eric Kever tenshi.sai...@gmail.com writes:

 I've created a file 'foo', and used tail -f to follow the changes to that
 file.
 I then wrote 'test' to the file and saved it, and tail reported 'test',
 which is fine.
 I then deleted 'test' and saved the file, and tail reported 'tail: foo:
 file truncated', which is fine.
 I then wrote 'test' again and saved the file, and tail reported 'est'
 instead of 'test'.

That's not a bug.  When you truncated the file you actually wrote a
single newline, so the current position became one character into the
file.  The fact that the next modification overwrote the newline (with
`t') wasn't noticed by tail, because it only watches for modifications
after the current end-of-file.

Try using an editor that actually allows you to write an empty file (or
use ` foo' in the shell).

Andreas.

-- 
Andreas Schwab, sch...@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
And now for something completely different.

bug#5941: tail bug

2010-04-13 Thread Pádraig Brady 
Tags: notabug