On Llu, 2003-06-23 at 01:46, David S. Miller wrote:

   From: Alan Cox [off-list ref]
   Date: 21 Jun 2003 13:36:54 +0100
   
   Standardising strings is a real help for end users,

I agree.  But my objections are in the context of doing this
inside the kernel, where such things do not belong.

Standardising strings for end users in the kernel is also good
because it both saves space and makes things more consistent for
the poor human wondering what blew up.

Standardising them for programs to parse is not a good idea