[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: Packaging guidelines: IPv6



On Wed, Jul 05, 2006 at 12:43:26PM +0100, David Woodhouse wrote:
> Since we're going through all Fedora packages (well, at least Core for
> now; I'm sure we'll get to Extras next) and filing bugs for any which
> lack IPv6 support, should we also update the guidelines on the Wiki to
> recommend that any package which supports Legacy IP should also be able
> to use IPv6 and should do so 'out of the box' where appropriate?
> 
> Would anyone object if I amended the PackageReviewGuidelines to include
> something along the lines of...
> 
> SHOULD: If any form of networking over IPv4 networking is supported, the
> same functionality over IPv6 should also be supported, and should be
> enabled by default if the IPv4 support is.
> 
> MUST: If IPv4 networking is supported, but for some reason the 'SHOULD
> support IPv6' documented elsewhere is not obeyed, a bug must be opened
> which should block the IPv6 tracker bug, and should contain a full
> justification for the lack. 

  My experience with IPv6: 3 years of portability nightmare since I accepted
the patch provided by Sun to 'add IPv6 support' in libxml2. That single
contribution led to most of the portability problems on verious unixes,
different breakages on different version of AIX, troubles with HP-UX and
don't ever start to look at the various Windows targetting toolchains.

  So my answer to your MUST is:
    give me configure and code patches which won't break on the various
    legacy platforms I try to support in my software.

My experience with it as libxml2 upstream maintainer has been abysmal, maybe
it's due to the quality of the initial patch, but what you try to present
as a done deal here, is still a not fully resolved issue upstream. Mandating
a fix here, is the equivalent of putting the blame on the people who want
to keep their software portable. If only the tone was cooled down that would
be easier to accept, but frankly your cruisade is not fun when you are stuck
on the other end.

  You want that feature, send the patches ! Pressuring the packagers
  or maintainers is just bad behaviour IMHO.

Daniel

-- 
Daniel Veillard      | Red Hat http://redhat.com/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]