Subject: Re: HEADS UP! Default value of ip6_v6only changed
To: None <tech-net@NetBSD.org, current-users@NetBSD.org>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
Date: 10/28/2003 06:38:50
> ok. so why this message in /usr/pkgsrc/www/MozillaFirebird-bin/MESSAGE?
> $ more MESSAGE.NetBSD
> Please note that if you have INET6 enabled in your kernel, you
> will need to set net.inet6.ip6.v6only=0, using sysctl(8).
Since I'm not responsible for any of the Mozilla code, much less the
networking code, I can only conjecture. My guess would be that the
author(s) was(were) lazy and, when v6 is available, insist on doing
everything, even v4 connections, through v6 sockets, instead of using
v4 sockets for v4 work and v6 sockets for v6 work - but couldn't be
bothered to explicitly set V6ONLY to 0 on the sockets in question.
This strikes me as incredibly lazy coding, but the only other ways I
cannot account for the message you quote are (1) that it's a historical
artifact which is no longer true, (2) the v6only sysctl actually
affects more than the documentation says it does, perhaps (for example)
disabling all use of v4-mapped addresses, even by sockets that have had
V6ONLY explicitly turned off, and (3) it was never true, but someone
mistakenly thought it was.
Actually, on reflection, I think alternative theory (1) isn't all that
much less plausible than the lazy-coding hypothesis.
> And why is it, that I cannot reliably browse into V6 nodes like the
> kame site to see the dancing turtle from this browser?
Again, I can only conjecture. This time my guess is that v6
connectivity between you and the site in question is not in good shape;
I know that in my own experience, v6 connectivity is often much worse
than v4 connectivity, even when the endpoint hosts are the same.
Comparing traceroute output with traceroute6 output may be useful.
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML firstname.lastname@example.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B