In your previous mail you wrote:
Should the MAY remain, with some discussion of why 3041 is good or should it
be upgrated to a 'SHOULD be implemented' with reasons why & why not to use it?
=> a SHOULD for implementation and a MAY for use should reach rough
consensus and the debate will
I agree, after re-reading the formal definitions in RFC 2119. And I
don't think we need any discussion text; the implications are already
covered in 3041.
The default can be left to the vendor IMHO.
Brian
Francis Dupont wrote:
>
> In your previous mail you wrote:
>
>Should the MAY rem
Hi Brian,
> I agree, after re-reading the formal definitions in RFC 2119. And I
> don't think we need any discussion text; the implications are already
> covered in 3041.
>
> The default can be left to the vendor IMHO.
I agree. I have added the text:
Privacy Extensions for Stateless Addres
On Tuesday, January 28, 2003, at 05:44 PM, [EMAIL PROTECTED]
wrote:
Privacy Extensions for Stateless Address Autoconfiguration
[RFC-3041]
SHOULD be supported. It is recommended that node behavior be
configurable
when they are available.
There is a catch to this last sentence.
Using
> > I agree, after re-reading the formal definitions in RFC 2119. And I
> > don't think we need any discussion text; the implications are already
> > covered in 3041.
> >
> > The default can be left to the vendor IMHO.
>
> I agree. I have added the text:
>
>Privacy Extensions for Statele
> Using RFC3041-type addresses is IMHO not a property of node,
> not a property of user context, and not even a property of applications.
> This is a property of the specific connections within each application.
I agree, privacy is the claim of user to determine for him or herself when
and how an
]; [EMAIL PROTECTED];
> [EMAIL PROTECTED]
> Subject: Re: Node Requirements and 3041
>
>
>
> On Tuesday, January 28, 2003, at 05:44 PM, [EMAIL PROTECTED]
> wrote:
> >
> >Privacy Extensions for Stateless Address Autoconfiguration
> > [RFC-3041]
> >
Hi Samita,
> In the last sentence, did you mean to say that the node should have a
> configuration knob whether to turn on privacy extension behavior ?
>
> If so, would it be clearer to say something like the following ?
>
> 'It is recommended that the node be configurable to turn on/off
> the
On Monday, February 3, 2003, at 05:28 AM, [EMAIL PROTECTED] wrote:
It should be something like that, but I think Alain was suggesting that
a node-based policy would not be sufficient. It should be at least
application based policy, since different applications will not
work with 3041 addresses.
Hi John,
> > If so, would it be clearer to say something like the following ?
> >
> > 'It is recommended that the node be configurable to turn on/off
> > the privacy extension for stateless address autoconfiguration, when
> > it is implemented.'
>
> It should be something like that, but I thi
Alain,
All your points are valid, but I'd argue the node requirements
document is the wrong place to have this level of detail.
Instead, 3041 should be respun with more text about these issues. I
think Node requirements would be better off (in general) just pointing
to other basic documents and l
: Loughney John (NRC/Helsinki); [EMAIL PROTECTED];
> [EMAIL PROTECTED]; [EMAIL PROTECTED]
> Subject: Re: Node Requirements and 3041
>
>
> Alain,
>
> All your points are valid, but I'd argue the node requirements
> document is the wrong place to have this level of detail.
&
By "toggle 3041 on and off" I mean decide if 3041 is appropriate or
not for that connection. Now, if the answer is yes, should a new address
be generated or should you use an existing one, this is, IMHO,
implementation dependent. I think it would make sense
to generate a new one per connection for
Alain,
By "toggle 3041 on and off", do you mean toggling address selection to use
an existing 3041 address, or generating a new 3041 address for each new
connection that wants it?
Brian
Alain Durand wrote:
>
> On Monday, February 3, 2003, at 05:28 AM, [EMAIL PROTECTED] wrote:
> >
> > It sho
14 matches
Mail list logo