Coreutils head and tail problems

[Daniel Berlin]

On Tuesday, August 26, 2003, at 1:33 PM, Phil Edwards wrote:

> On Tue, Aug 26, 2003 at 06:04:02PM +0100, Joseph S. Myers wrote:
>>
>> That simply means that replacement must be with more portable 
>> alternatives
>> such as "sed q" (as used in Paul Eggert's patch
>> <http://gcc.gnu.org/ml/gcc-bugs/2002-02/msg00516.html> to replace 
>> "head
>> -1").
> [...]
>> But I agree that dropping support for these options, absent
>> POSIXLY_CORRECT being set in the environment which GNU software would
>> traditionally have used for such a purpose, is ridiculously premature.
>
> You know, I'm usually the one arguing for adoption of new features.
> But the POSIX thought police can have my "uniq", "head -1", and "diff 
> -c"
> when they pry it from my Cold. Dead. Hands.
Seriously.
It's also a bit ironic that we now have to test to make sure the 
portable utilities we are using in the build system work with various 
options on each platform, when THAT'S THE REASON WE USED THEM IN THE 
FIRST PLACE (ie because they were portable and we didn't have to deal 
with various syntaxes).
Like, duh?

--Dan