Re: [Sks-devel] [CONTRIB] sks.pod text for enabling IPv6

[Phil Pennock]

On 2009-03-30 at 07:38 +0000, Kim Minh Kaplan wrote:
> Phil Pennock writes:
> > =head1 ENABLING IPv6
> >
> > This is more complicated than it should be, thanks to some limitations
> > in the language support libraries of O'Caml.  Instead of just being able
> > to do The Right Thing, you need to both explicitly enable IPv6 and
> > perhaps to understand a little about how your system deals with IPv4 and
> > IPv6 connections to an IPv6 socket.
> 
> I think this ranting should not go into the documentation.

*shrug*  You see it as a rant.  It's not, it's a plain explanation.

"Why is X more complicated when done by A than it is when done by B, C
or D?" -- a reasonable question, given that most modern software can
just do The Right Thing, thanks to AI_ADDRCONFIG and AI_V4MAPPED for
getaddrinfo() and IPPROTO_IPV6:IPV6_V6ONLY for setsockopt().

So the answer is "Because of limitations in the language support
libraries of O'Caml".  That's not a rant, just the truth.

As a sysadmin, when I'm asking "Why on earth is it doing something so
silly?" I appreciate honest answers rather than circumlocution.  I can
accept support library limitations as a reasonable answer and then do
things normally.  Without any kind of answer, I start wondering what I'm
missing and will get side-tracked trying to chase down what I'm not
understanding in the model.  I don't expect all languages to have
perfect support libraries -- I haven't seen a language yet which does.

If I cared more for the O'Caml language than I do, I'd look into
contributing back patches to the language maintainers to solve this
problem.  But that doesn't solve the immediate problem, unless an SKS
release will be delayed pending a new language compiler release and time
for everyone to move across to it.

But hey, you asked for documentation examples, I provided my suggested
text, but I don't really care if it's used or not.

-Phil

pgpcSGnrERsDv.pgp
Description: PGP signature