Posix2 conformance (was Re: coreutils - tail -f broken ???)

Matthew Burgess ca9mbu at hermes.sunderland.ac.uk
Thu May 8 02:27:46 PDT 2003


On Thu, 8 May 2003 05:35:11 +0000 (UTC)
Ryan.Oliver at pha.com.au wrote:

> So the way I see it we have a few options
> 1) build coreutils overriding the _POSIX2_VERSION supplied in
> posix/unistd.h
>    and build them non-compliant
> 2) edit the posix/unistd.h header and build everything non-compliant
> 3) grin and bear it, fixing up bugs and forwarding them on to the
> upstream
>    package maintainers
>    Means we have to grep for head/tail/sort etc through everything and
>    validate the usage
> 
> 2 is the easiest, 1 is most localised, 3 probably the most socially
> responsible.
> 
> Thoughts?

The first thought I had on this was to go with a combination of 2 and 3. 
i.e. The instructions in the book edit the unistd.h header so as failures
are not seen by those using the book.  However, those of us doing active
development do a few builds without changing the header file and report
bugs/fixes back to the upstream package maintainers.  Once the fixes are
available in upgraded package tarballs then the book should no longer
have to edit the header file (although a warning about non-standard
conforming shell scripts may be appropriate).  In my mind this is a
sensible approach as it allows for a somewhat smoother migration, and of
course means that LFS doesn't become the path-fest that it might have to
if it attemptes to be completely standards conformant straight off the
bat.

Just my tuppence.

Matt.
-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message



More information about the lfs-dev mailing list