Change of historical behavior

[Dave Jones]

On Sun, Apr 29, 2007 at 06:39:01PM -0500, Tom spot Callaway wrote:
 > On Sun, 2007-04-29 at 13:29 -0600, Dax Kelson wrote:
 > > ...... Original Message .......
 > > On Sun, 29 Apr 2007 08:25:29 -0500 Tom "spot" Callaway 
 > > <tcallawa at redhat.com> wrote:
 > > >Not to ask stupid questions, but should that have ever worked? I'd think
 > > >an outright rejection would be far useful than creating empty files
 > > >innocently.
 > > >
 > > 
 > > It should work because inodes and data blocks are two distinct resources 
 > > within a filesystem that UNIX/Linux quotas have the flexibility in 
 > > controlling separately. 
 > > 
 > > Behaviors (especially those with decades long precedence) shouldn't change 
 > > without very good reason, and they certaintly shouldn't change 
 > > non-deliberately.
 > > 
 > > I wonder if POSIX has anything to say about quota behavior.
 > > 
 > > Since I posted the message, I checked a SLES10 box using ext3 with acls and 
 > > user_xattr. It's behaves as expected in the traditional manner. Being over 
 > > your data block quota doesn't impact the ability to create zero byte files.
 > 
 > This may be a legit bug then. Bugzilla it. :)

It's nothing we added. Tell upstream would be a better answer.

	Dave

-- 
http://www.codemonkey.org.uk