On 11/05/2012 03:56 PM, Ganton wrote:
>> Sorry, but dd is older than POSIX
> Paul Eggert wrote "dd is [...] part of the POSIX standard" and I wrote 
> consequently, if the dd specification is broken, then the POSIX standard is 
> broken, too.
> 
>> You're 40 years too late on this one.
> This kind of condescending attitudes do not improve matters.

I'm sorry if I came across as condescending - that was not my intent.
Email is a poor forum for hearing the emotions intended.

But the point remains - if you think POSIX is broken for having
standardized dd behavior (40 years old) that existed at the time POSIX
was first written (20 years old), then please take up your complaint
with POSIX.  Membership in the Austin Group is free of charge:
http://www.opengroup.org/austin/.  I will give you fair warning in
advance that you are unlikely to convince the Austin Group to make any
changes, because they are reluctant to change historical behavior that
has been so firmly entrenched for so many years, but more power to you
if you succeed.  Remember, dd is already an oddball for taking arguments
in the form of 'param=val' instead of the more traditional '-<letter>
val', precisely because dd is so much older than the bulk of the other
utilities that were standardized by POSIX; that is, dd existed prior to
the common practice of being silent on success that exists in many of
the other utilities also standardized by POSIX.

Meanwhile, Coreutils can't do anything about the default situation
unless you can convince the POSIX people to change the standard; and
likewise, we can, and have, done something about it if you are willing
to use extensions beyond POSIX, in the form of status=none.

> 
>> No need to report a new bug - we already told you that coreutils 8.20
>> added 'dd status=none' to silence even that information.
> "We already told you"? Who is Eric Blake talking to? Nobody talked there about 
> coreutils 8.20 or 'dd status=none'.

Pádraig mentioned 'dd status=none' in the message before your reply
addressed to Paul, when looking at the thread:
https://lists.gnu.org/archive/html/bug-coreutils/2012-11/msg00012.html
or at the bug report:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12794#13

> Paul Eggert kindly wrote about 'dd 
> status=noxfer'. Eric Blake can re-read the thread if he wants to, and if he 
> doesn't want to, at least he can stop talking people condescendingly after not 
> really reading the conversations.

Again, I apologize if you are mis-reading my intent.  And yes, I did
re-read the thread, both before my first thread mentioning why POSIX
standardized things the way they did (and therefore why coreutils can
only add extensions, such as status=none, rather than changing the
behavior by default), and again before writing this reply; my assumption
(perhaps mistaken) was that you had read Pádraig's reply.

-- 
Eric Blake   eblake@redhat.com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org