Re: [v6tc] Re: Tunneling and Transition Drafts

[Jeroen Massar <[e-mail]>]

On Fri, 2005-04-08 at 12:42 -0400, W. Mark Townsley wrote: 

<SNIP>

>  > When used in this mode then it will,
> > just like the others cross a NAT with ease, unless firewalled of course.
> > Because of the Keepalive protocol it will also stay 'up'. When the
> > source address changes (dhcp change, change of wireless net) one will
> > also automatically reconnect, might have a small delay but should work.
> > 
> > We still miss an actual implementation on the more widely deployed
> > platforms and as those are in the big target area this will be something
> > many people will want. I am actually tempted all of a sudden to just
> > implement it, but unfortunately not enough time for that.
> 
> Or be tempted to implement IPv6 for PPP on Windows XP and sell it (or offer it 
> as shareware) for ~$9.95 for a while... Until Longhorn comes out, you might make 
> a little money on the side... (whether L2TP is used for v6tc or not).

Asking money for a transition tool is silly. Helping people get it for free
and with ease is my target. There is, at the moment, IMHO, no big point in
demanding money for such services. And especially with the 'free copying'
going on in this world it doesn't make sense anyway unless you are able to
afford time and money on reclaiming it, which simply is not possible anyway.
For these cases, just let people donate if they need to. Anyway this is the IETF ;)

Nevertheless I'll put it somewhere on that big list, lets see how many parties
get canceled this week...

> In the end, a lot of implementations may actually be CPEs rather than hosts, in 
> which case having the linux implementation helps.

Linux supports it, but not too many people want to actually touch their gear.
and for most of those people Linux is still scary.

On Fri, 2005-04-08 at 19:32 +0200, Francis Dupont wrote: 
>  In your previous mail you wrote:
> 
>    L2TP ?
>    
> => L2TP works in very bad networks (private addressing, paranoid filtering,
> etc) so IMHO we should keep it in our toolkit. It is a standard,
> there are many implementations (I use it over every Linuxes and BSDs)
> and it is ISP friendly (ISPs are still fond of PPP :-).
> 
> Regards
> 
> [e-mail]
> 
> PS: it seems a cleaned TSP and L2TP are enough for v6tc stuff.

With, IMHO an addition of a discovery protocol.

Do note btw that TSP is a configuration protocol, it does not cover the
NAT-crossing case.

On Fri, 2005-04-08 at 20:06 +0200, Francis Dupont wrote:
>  In your previous mail you wrote:
> 
>    Since when does proto-115 cross NAT's ?
>    
>    L2TP typically does not work for the same reason why proto-41 does not
>    work: NAT boxes don't forward proto-41 nor proto-115.
>    
> => usually L2TP is run over UDP, not protocol 115.

As was indeed pointed out already ;)

>    Also L2TP has the limitation that it requires authenticated use, which
> 
> => I don't understand your concern: is it about the optional shared secret
> to protect L2TP control or PPP authentication?

I actually like authentication, as that is a good thing, because it
allows also to track abuse easier. But I recall that there where a
number of scenarios where people demanded 'anonymous tunneling'.

>    My take on this thus becomes:
>    - In general tunneling IPv6 over UDP works *everywhere*
>      (unless there is an administrative filter prohibiting it
>       in which case we should not try to override it either)
> 
> => there are two very different instances of this: Teredo and L2TP.
> I don't believe we need a third one.

Teredo could be used for the case where the ISP doesn't supply it, L2TP
for the case where they do.

>    - TSP is deployed and works (*)
>    
> => yeah! TSP is great in the (common) cases you can use it. Perhaps too
> great so there are not enough TSP services where I am...

Off-list/topic/subject:

Afaik you are in France. Should I enable the TSP server for the
SixXS system? Then you have near-European wide coverage...
If someone wants to push me over the convince border, yell privately.

Next to that, you are in an french educational network, thus I guess
that Jerome Durand / Renater can supply you with such connectivity.

Also makes the claim that "TSP" is not deployed go away.

> PS: I don't believe we have to spend a lot of time on marginal improvements:
> the ultimate target should to get *native* IPv6!

My idea ;) But unfortunately not always possible.

Greets,
 Jeroen

_______________________________________________
v6tc mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/v6tc