> Please read my summary of the situation. Debian has been shipping its
> own md5sum since at least in August 1994. It's the GNU people who've
> made a change to the behaviour.
Yes, and I think that the time to lobby GNU to conform to the Debian
md5sum behavior was long before GNU's md5sum established itself as THE
md5sum. No one uses PGP2's md5sum, and I certainly wouldn't write a
script depending on the "Debian behavior".
> The time for changing Debian md5sum's behaviour has long passed.
Perhaps, but there's still time to retire it.
> Also, you don't address my argument that the original behaviour is
> much more convenient.
I don't think it's relevant. However, since you seem to, I will point
out that it's not inconvenient for me at all.
print ${"$(print test | md5sum)"%% *}
will give me the same result whether md5sum is Debian's historical
md5sum, Debian's "rogue" md5sum, GNU's md5sum, or an alias to *BSD's md5.
Furthermore, if someone packages a BSD md5, you can get the dashless
output merely by piping to md5(1), which is 3 keystrokes more
convenient.
