[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Regression in svnserve: IPv6 causes IPv4 not to work [was: Branching for 1.3.]

From: Sigfred Håversen <bsdlist_at_mumak.com>
Date: 2005-09-23 23:40:44 CEST

Greg Hudson wrote:
> On Fri, 2005-09-23 at 08:18 +0200, Sigfred Håversen wrote:
>
>>Attached a patch for the --ipv4-only variant. On OpenBSD
>>you can run two copies of svnserve, one for IPv4 and one for IPv6.
>
>
> Why can't we just create two listening sockets, one for IPv6 and IPv4?
>
> (If the only issue is implementation difficulty, I'm -1 on adding a new
> option out of laziness, since we'd have to support it until 2.0.
> Ripping out IPv6 support would be a better option from my perspective.)

If you want more than one socket I see no reason not to support n>1 sockets,
and give users the choice of using more than one --listen-host as well. But
this is not what my posts has been about.

My suggestion, like the the poster in May, was quite simply to have an
option whether or not to use IPv6 as a user option; not to have it
forced upon us with no choice what so ever. Removing IPv6 and leaving
existing IPv6 users in the dark was certainly not in the agenda for
my part.

My quote from an earlier post was to point out that an IPv6 imlementation
for svnserve is not that straightforward if you want to have it portable.
It did not imply any kind of laziness of the implementors of said
functionality. In case, removing IPv6 is out of the question since you'll
have to support it to 2.0, at least ;-)

/Sigfred

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Sep 23 23:41:30 2005

This is an archived mail posted to the Subversion Dev mailing list.