Re: Something better than DNS?

[Douglas Otis]

On Nov 29, 2006, at 8:53 AM, Hallam-Baker, Phillip wrote:


I don't think that would be the only patent you would need 


Here is a somewhat more complete list:

http://ops.ietf.org/lists/namedroppers/namedroppers.2006/msg01076.html

-Doug



___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

RE: Something better than DNS?

[Hallam-Baker, Phillip]

> From: Edward Lewis [mailto:[EMAIL PROTECTED] 
> At 11:42 -0500 11/29/06, Emin Gun Sirer wrote:
> 
> >Let's not torque the discussion off topic. Free market 
> economics does 
> >not come to bear on the issue because there is no free 
> market to speak 
> >of for registries.
> 
> What did I say about frictionless surfaces?

Edward, assume frictionless surfaces is what a physicist says to an engineer. 
This is economics, a consensual delusion on a much grander scale.


We did not merely assume the surface to be frictionless: by making the 
assumption we caused the surface to become frictionless.

Having assumed a free market we have caused the market to become free of all 
defects such as time delay, liquidity, imperfect knowledge and all the other 
real world effects without which Wall Street would make no money.

Having assumed that people behave as perfect evaluators of their narrow self 
interest and act in strict accord with those interests we have turned the 
entire population into selfish bastards who can calculate their best interest 
to three decimal places.



___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: DNS Choices

[Dave Crocker]

[EMAIL PROTECTED] wrote:

That's why it seems quite reasonable to continue
work on the protocol as a more general distributed
database service. But not on port 53 which is
mission critical for the ONE NETWORK which
rules them all.



The essential argument that you are making, I think, is about the DNS
*operational* service, rather than about the DNS protocol.  Hence the call for
using a different port for additional uses. (cf., SMTP and SUBMISSION, of 
course.)

Yet responses so far are about the protocol, or maybe about protocol "support".

The observation that the vast bulk of current DNS use is for a particular
function that is essential to real-world use of the basic infrastructure ought
to be obviously true to folks.

So the question is whether adding other uses of it might pose any significant
problems.  This is not a protocol question, but an ops, admin, and management
question.

Will the different uses create traffic patterns, administrative requirements, or 
the like, that threaten the current, essential service, in any of the ways 
previously cited?


A small example:  The current DNS is typically administered by a particular kind
of group within an organization's IT structure.  Do the proposed new uses pose
any problems for this, by virtue of needing a possibly different set of
administrators?  (It turns out that the controversial underscore naming
technique provides a way of vectoring some new uses to new administrative 
groups.)


d/
--

  Dave Crocker
  Brandenburg InternetWorking
  bbiw.net



___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

RE: IM and Presence history

[Hallam-Baker, Phillip]

I absolutely agree with Steve here, but I think that the problem here is too 
little integration, not too much. I don't think that this security through 
obscurity scales very well.

There needs to be a gatekeeper. If someone wants to schedule a call with me, 
fine, just drop me a note first so I can tell my system to accept it. Oh and if 
you want to send more than a few lines in the note you will have to be on the 
approvals list.

CEOs and Paris Hilton already have these security measures in place.


I think that a good technical bar to set here is that a 'one address' system 
must be secure enough against unwanted contact that Paris Hilton can use it and 
post the same contact address on her Web site as Britney Spears would use to 
contact her.

If you are known directly or a friend of a trusted friend you get in, otherwise 
you get a lower level of communication, the bottom rank being directed to the 
Paris Hilton fan club.


> -Original Message-
> From: Steven M. Bellovin [mailto:[EMAIL PROTECTED] 
> Sent: Wednesday, November 29, 2006 7:23 PM
> To: [EMAIL PROTECTED]
> Cc: Harald Alvestrand; ietf@ietf.org
> Subject: Re: IM and Presence history
> 
> On Wed, 29 Nov 2006 10:33:15 -0800
> Dave Crocker <[EMAIL PROTECTED]> wrote:
> 
> 
> > 
> >   The underlying specifications permit you to have different 
> > addresses, for different services.  They also permit you to have the
> > *same* address.
> > 
> This is only a good idea if coupled with a powerful, 
> easy-to-use interface that restricts presence visibility.  
> Many more people have my email address than my IM addresses; 
> I'm also very careful about who gets my mobile phone number.  
> Why?  Because IM communication and phone calls interrupt me 
> in a way that email does not.  In fact, I take advantage of 
> email to avoid giving out the other information promiscuously 
> -- I tell people who perceive an urgent need to reach me to 
> email page-smb at the appropriate domain; this address is 
> translated to both SMS and a direct email message.
> 
> 
> 
>   --Steve Bellovin, http://www.cs.columbia.edu/~smb
> 
> ___
> Ietf mailing list
> Ietf@ietf.org
> https://www1.ietf.org/mailman/listinfo/ietf
> 
> 

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: IM and Presence history

[Steven M. Bellovin]

On Wed, 29 Nov 2006 10:33:15 -0800
Dave Crocker <[EMAIL PROTECTED]> wrote:


> 
>   The underlying specifications permit you to have different
> addresses, for different services.  They also permit you to have the
> *same* address.
> 
This is only a good idea if coupled with a powerful, easy-to-use
interface that restricts presence visibility.  Many more people have my
email address than my IM addresses; I'm also very careful about who
gets my mobile phone number.  Why?  Because IM communication and phone
calls interrupt me in a way that email does not.  In fact, I take
advantage of email to avoid giving out the other information
promiscuously -- I tell people who perceive an urgent need to reach me
to email page-smb at the appropriate domain; this address is translated
to both SMS and a direct email message.



--Steve Bellovin, http://www.cs.columbia.edu/~smb

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: IM and Presence history

[Dave Crocker]

Henning,


Henning Schulzrinne wrote:
you might want to look at the SIP design, which offers most of the 
functionality you describe already. The notion of a common address 
(prefixed to generate a URL by the communication scheme, be it sip: or, 
more generically, pres: or im:) were part of the design, 



SIP obtained this design from previous work on IM and Presence:

 
contains the announcement.


I can't seem to track down the original draft of draft-ietf-impp-srv-00, 
"Address Resolution for Instant Messaging and Presence" which provided the 
initial generalization.



d/
--

  Dave Crocker
  Brandenburg InternetWorking
  bbiw.net

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Carl Malamud]

Hi -

I actually think the question of how a namespace is to be administered
is a perfectly valid one for the IETF to consider if it impacts the 
performance or functionality of a protocol.  

We do that all the time when we give explicit instructions to the IANA 
in an "IANA Considerations" section.  The IETF could also do the same 
in an "ICANN Considerations" section when it comes to the DNS.  

Carl

> Brian E Carpenter wrote, On 29/11/2006 10:43:
> > your question is linked to whether we treat the namespace as a public
> > good to be administered for the greater public good, or as a
> > commodity to be treated like coffee beans. And that really isn't
> > a question for this technological community.
> Depends how you look at it. Technological choices do have an impact on
> what the society is able to do (or not). It is up to the society to tell
> the engineers what it needs/wants. But I agree there are other fora for
> this.
> 
> Patrick
> 
> ___
> Ietf mailing list
> Ietf@ietf.org
> https://www1.ietf.org/mailman/listinfo/ietf
> 

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Stephane Bortzmeyer]

On Wed, Nov 29, 2006 at 12:40:09PM -0500,
 Edward Lewis <[EMAIL PROTECTED]> wrote 
 a message of 56 lines which said:

> The terminology used here indicates a need for a deeper understanding of 
> DNS.

I suspect that he is deliberately trolling, in order to prove a point
(that DNS is too limited to handle domains who need a lot of
reliability).


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Stephane Bortzmeyer]

On Wed, Nov 29, 2006 at 11:42:17AM -0500,
 Emin Gun Sirer <[EMAIL PROTECTED]> wrote 
 a message of 36 lines which said:

> Right now, we'd like to have a domain delegated to a large number
> (say 100+) of nameservers.

See Edward Lewis' respond (basically, global anycast + local anycast
and you have as many servers as you wish).

> The registrars we have gone through impose a limit on the number of
> nameservers they are willing to accept.

The limit typically comes from the registry (eight nameservers in
".fr").

> (And no, not all nameservers need to be returned in response to
> every query. A random sampling would be fine).

Since they are all part of the same RRset, it is not legal to return
only a part of them. A RRset must be returned completely (or the TC
bit set).

BTW, did you test if typical resolvers (say, BIND) can handle as many
nameservers? I doubt it, they probably have a fixed array to store the
nameservers they try.

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: The IESG Approved the Expansion of the AS Number Registry

[william(at)elan.net]

On Wed, 29 Nov 2006, Joe Abley wrote:


I did not see any consensus on that issue when it was brought to NANOG-m.


Interesting. I didn't notice any support for separating the 32-bit quantity 
into two sections, but I remember many people decrying the need for any 
separator at all.


I'd have to look at archive but I was not left with impression that people 
were against it rather they were noting changes that would have to be made.


The principal argument against "." specifically was that it will instantly 
break all deployed AS_PATH regular expressions (or at least, potentially 
cause regex comparisons to provide surprising results).


I'm pretty sure it would be some time before tools are available that will
provide AS_PATH in new format and even then I've a feeling same tools will
give user an option to get data with ASN in new format or get it with ASN 
as a 64-bit number. Note that large number of existing ASN-processing 
scripts would break any way you go if they are doing bound checking...


--
William Leibzon
Elan Networks
[EMAIL PROTECTED]

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: The IESG Approved the Expansion of the AS Number Registry

[Joe Abley]

On 29-Nov-2006, at 12:14, william(at)elan.net wrote:


On Wed, 29 Nov 2006, Joe Abley wrote:


On 29-Nov-2006, at 08:30, Henk Uijterwaal wrote:


[EMAIL PROTECTED] wrote:

On the NANOG list it has already been pointed out that a lot
of network management software cannot handle such notation and
in some cases, 1.0 could be interpreted as the IP address  
1.0.0.0. It has been confirmed that one widely used PERL library  
interprets x.y as IP address x.0.0.y.

I think this is a bug.


If it is, it's a very long-standing one. For example, see INET(3)  
which I think is of 4.2BSD vintage, and which appears to have  
similar semantics to the mentioned perl library:


I fail to see that as being stopping point.


I didn't suggest it was. I was just responding to the thought that  
the interpretation of x.y as x.0.0.y has some history.



The draft above received significant operator criticism.

The consensus I saw on NANOGm, for example, was that there was (a)  
no useful reason to be able to distinguish between a 16-bit AS  
number and a 32-bit AS number less than 65536, (b) no good reason  
to use punctuation to separate the most- and least-significant 16  
bits of the 32-bit ASN, and (c) every reason to think that the  
most sensible representation was just "bigger decimal numbers".


I did not see any consensus on that issue when it was brought to  
NANOG-m.


Interesting. I didn't notice any support for separating the 32-bit  
quantity into two sections, but I remember many people decrying the  
need for any separator at all.


The principal argument against "." specifically was that it will  
instantly break all deployed AS_PATH regular expressions (or at  
least, potentially cause regex comparisons to provide surprising  
results).


I'm not making an argument for against ggm's draft (although I can if  
that seems useful :-) I was merely passing on my memory of the NANOG  
thread.



Joe


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: IM and Presence history

[Henning Schulzrinne]

See 
http://www.softarmor.com/wgdb/docs/draft-schulzrinne-sipping-id-relationships-00.txt 
for an expired draft on this topic.


There is an architectural 'trick' here, that I suspect is the key for 
making thing homogenize in a way that is tractable:


 The underlying specifications permit you to have different 
addresses, for different services.  They also permit you to have the 
*same* address.


So the fact that your jabber and email and... (whatever) services all 
get data to you via "[EMAIL PROTECTED]" is an administrative choice, 
not one imposed by some grand unifying architecture that needed to be 
designed perfectly from the start.


The only "architectural" rule needed for this is to recommend that folks 
base new adddressing on an existing scheme, to avoid collissions.  For 
example, an administrative rule that foo:[EMAIL PROTECTED] is only 
available for registration to the recipient of 
mailto:[EMAIL PROTECTED] is all that is needed to make this work.


(Anyone paying close attention will note that this introduces a problem 
with getting a foo: address that is not the same as the email address 
but is not assigned to anyone else. But what the heck, I'm not trying to 
design the whole thing right now...)


At any rate, this is a version of the "think globally, act locally" 
approach to architecture design that good Internet technical work did well.


d/



___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: IM and Presence history

[Dave Crocker]

Janet P Gunn wrote:

The original Ethernet? (not really "discontinuous", but quite a big leap)


I think that Ethernet, like the Web, are actually excellent COUNTER-examples. 
Ethernet is Alohanet with carrier-sense, collision-detect, exponential backoff. 
 And, of course, it runs over wire rather than radio.


IMO, Ethernet and the Web represent (high quality) incremental work that had 
discontiguous impact.


The major part of the incremental work was systems-level synthesis of just the 
right set of features, at the right time.


d/

--

  Dave Crocker
  Brandenburg InternetWorking
  bbiw.net

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: IM and Presence history

[Dave Crocker]

Harald Alvestrand wrote:
one nice thing about the schema/protocol being part of the naming scheme 
is that it does *not* tie me to a single provider for all services - my 
jabber service for [EMAIL PROTECTED] is provisioned from someone 
who's got no relationship at all to my mail and web services.



There is an architectural 'trick' here, that I suspect is the key for making 
thing homogenize in a way that is tractable:


 The underlying specifications permit you to have different addresses, for 
different services.  They also permit you to have the *same* address.


So the fact that your jabber and email and... (whatever) services all get data 
to you via "[EMAIL PROTECTED]" is an administrative choice, not one imposed 
by some grand unifying architecture that needed to be designed perfectly from 
the start.


The only "architectural" rule needed for this is to recommend that folks base 
new adddressing on an existing scheme, to avoid collissions.  For example, an 
administrative rule that foo:[EMAIL PROTECTED] is only available for 
registration to the recipient of mailto:[EMAIL PROTECTED] is all that is 
needed to make this work.


(Anyone paying close attention will note that this introduces a problem with 
getting a foo: address that is not the same as the email address but is not 
assigned to anyone else. But what the heck, I'm not trying to design the whole 
thing right now...)


At any rate, this is a version of the "think globally, act locally" approach to 
architecture design that good Internet technical work did well.


d/

--

  Dave Crocker
  Brandenburg InternetWorking
  bbiw.net

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Fwd: The IESG Approved the Expansion of the AS Number Registry

[Geoff Huston]

But, with the expanded space, there is an issue of how to transition
to the larger numbers. This is a software problem as much as
anything. Until all software understands the bigger numbers, people
will want to continue using the 16-bit ones.



I had a shot at documenting this in the form of a presentation slide 
pack. Even if you continue to use a 16 bit ASN indefinitely things 
may need to change:


http://www.potaroo.net/presentations/2006-10-11-asns.pdf





Meanwhile, to encourage the migration to 4-byte ASNs, the RIRs have
been developing a globally-coordinated policy for how to transition to
4-byte ASNs (hi Geoff!). The first step of that process, effective
January 1, 2007, is to allow requestors to obtain 4-byte ASNs (if they
specifically request them). But, in order for the RIRs to be able to
offer them, IANA has to give them some. Hence, the IESG action
mentioned in the note.

See http://www.arin.net/policy/proposals/2005_9.html for more details.



and http://www.potaroo.net/ispcol/2005-08/as.html for a (much longer) 
look at the role of ASNs and the way in which expanded AS numbers are 
to be transitioned in from a technical perspective


regards,

   Geoff




___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Edward Lewis]

At 11:42 -0500 11/29/06, Emin Gun Sirer wrote:


Let's not torque the discussion off topic. Free market economics does
not come to bear on the issue because there is no free market to speak
of for registries.


What did I say about frictionless surfaces?


A quick question: Right now, we'd like to have a domain delegated to a
large number (say 100+) of nameservers. The registrars we have gone


The terminology used here indicates a need for a deeper understanding of DNS.


through impose a limit on the number of nameservers they are willing to
accept. Is this a limit stemming from the .COM registry, or is there a
registrar out there that will let us delegate a .COM domain to a few
hundred nameservers? (And no, not all nameservers need to be returned in
response to every query. A random sampling would be fine).


The conventional limit on the number of name servers for a zone is 
derived from the pre-EDNS0 limit of 512 bytes in a UDP'd DNS message. 
When you query for the root zone SOA you get back the SOA, 13 NS 
records, and 13 glue A records.  If it were 14, then the message 
would be truncated and probably lead to a TCP connection attempt.


From this hard (operations) limit, the notion of 13 as maximum crept 
into registry software.  And not just .COM.


Ideally, the limit would be lower, as IPv6 wasn't invented when the 
number 13 was derived.  To stay under the 512 limit and still include 
 records, a lower limit would be needed.


Your parenthetical comment is contrary to one of the most important 
principles in the DNS, coherency.  Especially high up in the 
hierarchy.  It would be hard to debug problems if the returned set of 
servers in a referral changed very rapidly.  There is the practice of 
tailoring answers to a querier, although this is globally incoherent 
in the strict sense at least the answers to the querier in question 
stay fairly constant (and coherent from different available sources).


Of course, the 13 name limit does not limit you to 13 name servers. 
With anycast, the number can be unbounded subject to the concerns 
with routing.  And if you add in load balancers you can have even 
more servers.  It depends on how you count 'em.


--
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Edward Lewis+1-571-434-5468
NeuStar

Dessert - aka Service Pack 1 for lunch.

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: The IESG Approved the Expansion of the AS Number Registry

[william(at)elan.net]

On Wed, 29 Nov 2006, Joe Abley wrote:


On 29-Nov-2006, at 08:30, Henk Uijterwaal wrote:


[EMAIL PROTECTED] wrote:


On the NANOG list it has already been pointed out that a lot
of network management software cannot handle such notation and
in some cases, 1.0 could be interpreted as the IP address 1.0.0.0. It has 
been confirmed that one widely used PERL library interprets x.y as IP 
address x.0.0.y.


I think this is a bug.


If it is, it's a very long-standing one. For example, see INET(3) which I 
think is of 4.2BSD vintage, and which appears to have similar semantics to 
the mentioned perl library:


I fail to see that as being stopping point. ASN is not an address and 
should not be passed to INET library - after all you do not pass just

a number to the ASN-library just because its a number? Also note that
for ASNs the bounds for A.B parts (0-65535) are completely different then 
A.B.C.D for ip address (0-255).


BTW - 3 or more "." is also commonly used in representing phone numbers
and I'm sure other things too.


The draft above received significant operator criticism.

The consensus I saw on NANOGm, for example, was that there was (a) no useful 
reason to be able to distinguish between a 16-bit AS number and a 32-bit AS 
number less than 65536, (b) no good reason to use punctuation to separate the 
most- and least-significant 16 bits of the 32-bit ASN, and (c) every reason 
to think that the most sensible representation was just "bigger decimal 
numbers".


I did not see any consensus on that issue when it was brought to NANOG-m.
There was critisicm but its not anywhere near that majority said this 
notation is bad - in fact I think its the other way around and most 
thought it was fine.


As far as reasons:
 1. We do not use full 32-bit number when talking about ip address for
good reasons - it would make things more difficult for humans who
need to remember and communicate them [ok - there are other historic
rasons too and CIDR based use as well...]. We put "-" and "." for
phone numbers eventhough there its not like 32-bit number and its
all same digits no matter with or without "-". Its all largelye due
to that large numbers are not natual to humans... So as number like
18.101 is easier then 1179749 and that helps when you're doing
manual debugging.
 2. There is are reasons to believe that ASNs can be assigned so that
in new notation 2.x would indicate its ARIN and 3.x is RIPE region.
Its not necesarrily relevent for every case, but it does help.
 3. Several different punctuation marks were explored on several lists
including ppml - generally people thought that "." was easiest due
to its use in ip addresses. You seem to argue the other way around...

--
William Leibzon
Elan Networks
[EMAIL PROTECTED]

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

RE: IM and Presence history

[Hallam-Baker, Phillip]

> From: Brian Rosen [mailto:[EMAIL PROTECTED] 

> > However, what this subthread demonstrates is that they were 
> > conceptually an incremental change, not a giant, discontinuous, 
> > intellectual leap.
> > 
> > I thought we all knew that.
> Oh, I agree, we knew that.   There are very, very few discontinuous
> intellectual leaps in our part of the universe.  It's hard to 
> name one that happened in the past decade or two.
> 
> Can we name any discontinuous intellectual leaps of late in 
> computer networking?  Now that I think about it, forget "of 
> late", have there EVER been any?

People not bits: The idea that computing is about the user, not the machine.

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

RE: SMTP compared to IM (Re: DNS Choices: Was: [ietf-dkim] Re: Last Call: 'DomainKeys)

[Hallam-Baker, Phillip]

Before the web it was possible to be on a different network and still exchange 
email. It did not work at all well but it did work sorta.

Even though the web did in theory work on other protocols (I ran a server on 
HEPNET) most of the content was on the Internet.

So there was a different value proposition when someone proposed getting an 
Internet connection. People could no longer be fobbed off with 'The JANET 
gateway already allows you to exchange mail with the Internet, its cloured 
books for you until we deploy OSI'.

> -Original Message-
> From: Eric Burger [mailto:[EMAIL PROTECTED] 
> Sent: Wednesday, November 29, 2006 8:06 AM
> To: John C Klensin; Dave Crocker; ietf@ietf.org
> Subject: Re: SMTP compared to IM (Re: DNS Choices: Was: 
> [ietf-dkim] Re: Last Call: 'DomainKeys)
> 
> Actually, as I fuzzily recall in the 1986 - 1992-ish period, 
> MCImail had a large presence for business messaging and 
> CompuServe had a lion's share of consumer messaging.
> 
> Before the flames go on, realize that (1) my memory is fuzzy 
> and (2) the market was seriously fractured.  The large 
> enterprise market was doing the Notes thing; the small 
> enterprise market was doing the cc:mail, netware, etc. thing, 
> and interoperability was something that people gave lip service to.
> 
> What a difference five years made!  By 1996, pretty much 
> everyone interoperated with Internet Mail.
> 
> 
> On 11/26/06 10:35 PM, "John C Klensin" <[EMAIL PROTECTED]> wrote:
> 
> > --On Friday, 24 November, 2006 10:30 -0500 Eric Burger 
> > <[EMAIL PROTECTED]> wrote:
> > 
> >> Or, the reality that with (at the time) a single dominant network 
> >> provider made the inter-networking point moot.
> > 
> > Eric, you are being a little cryptic, perhaps unintentionally.
> > What do you mean about a single dominant provider and at what time?
> > 
> > I would add an observation to Dave's about possibly 
> different sets of 
> > needs by reminding everyone that considerable IM 
> functionality (other 
> > than presence) isn't new.  We had SEND/SOML/SAML from the 
> beginning of 
> > SMTP, even though they had, IMO, a very short practical 
> lifespan and, 
> > even then, were used only in limited communities.  We also we had a 
> > couple of flavors of the "talk" protocol which were 
> certainly heavily 
> > used in some places.  "Talk" involved a conversational 
> session while 
> > SEND et al was closer to what we would call a short message service 
> > today.  Off the Internet and in the land of BITNET/EARN/etc., there 
> > was also an end to end short message protocol and mechanism 
> that was 
> > extensively used.
> > 
> > None of these supported a presence mechanism in the sense that we 
> > understand it today.  As a result, one had to bind a user 
> identity to 
> > a target host in much the way SMTP does, rather than having someone 
> > attach to the network at any point and announce presence and, 
> > implicitly, location.  It is arguably those presence and mobility 
> > mechanisms and not IM itself that is the recent 
> development.  To the 
> > degree to which those mechanisms are what caused IM to take off, 
> > perhaps that reinforces Dave's view of different services serving 
> > different needs.
> > 
> >  john
> > 
> >> On 11/22/06 11:13 AM, "Dave Crocker" <[EMAIL PROTECTED]> wrote:
> >> 
> >>> 
> >>> 
> >>> Harald Alvestrand wrote:
> > There were no alternatives to SMTP on an IP network 
> until Instant 
> > Messaging came along.
>  
>  not since X.400 over X.25 died, no. I thought you were 
> older than 
>  that
> >>> 
> >>> And there were all of the individual providers that 
> Michael cited, 
> >>> such as MCI Mail.
> >>> 
> >>> 
> >> but can be seen in IM, and may likely show up in other 
> forms of 
> >> communication.  Much of this is simply the nature of software.
> > 
> > It has nothing to do with software and everything to do with 
> > architecture. IM networks have less problems because all the 
> > participants share a relationship with the IM service providers.
> >>> 
> >>> It *is* interesting that the diversity of disconnected email 
> >>> services was viewed as a basic problem to solve, whereas 
> most of the 
> >>> Internet user community does not seem to feel the same 
> pressure to 
> >>> unify IM.
> >>> 
> >>> Hmmm.  Maybe IM satisfies a different set of needs than 
> does email.  
> >>> So we had better be a bit cautious about trying to generalize 
> >>> implications between them.
> >>> 
> >>> 
> >>> d/
> >> 
> >> 
> >> ___
> >> Ietf mailing list
> >> Ietf@ietf.org
> >> https://www1.ietf.org/mailman/listinfo/ietf
> > 
> > 
> > 
> > 
> > 
> 
> 
> 
> ___
> Ietf mailing list
> Ietf@ietf.org
> https://www1.ietf.org/mailman/listinfo/ietf
> 
> 

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/

RE: Something better than DNS?

[Hallam-Baker, Phillip]

I don't think that would be the only patent you would need  

> -Original Message-
> From: Douglas Otis [mailto:[EMAIL PROTECTED] 
> Sent: Tuesday, November 28, 2006 9:56 PM
> To: [EMAIL PROTECTED]
> Cc: Patrick Vande Walle; ietf@ietf.org
> Subject: Re: Something better than DNS?
> 
> 
> On Nov 28, 2006, at 4:31 PM, Emin Gun Sirer wrote:
> 
> > Stephane & Phillip,
> >
> > I'm thinking of writing a short report that summarizes the 
> invaluable 
> > discussion here and beefing up the system sketch. I think 
> we now agree 
> > that it is possible to have multiple operators manage names in a 
> > single, shared namespace without recourse to a centralized 
> > super-registry.
> 
> You might want to review patent 7,065,587 as well.  Rather 
> than a hierarchical name space, there are GUIDs and "friendly names"  
> socially structured into groups.  In addition to friendly 
> names, GUIDs can combine with DNS names as well.  There is no 
> need for a super-registry, but rather a way to generate 
> GUIDs.  Perhaps this is the structure of things to come, 
> where belonging to a group matters more than a centralized authority.
> 
> -Doug
> 
> 
> ___
> Ietf mailing list
> Ietf@ietf.org
> https://www1.ietf.org/mailman/listinfo/ietf
> 
> 

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

RE: Something better than DNS?

[Hallam-Baker, Phillip]

> From: Harald Alvestrand [mailto:[EMAIL PROTECTED] 

> Emin Gun Sirer wrote:
> >
> > As an Internet user, I wonder about two things in the long term:
> > - why is it so expensive to register a name?
> > - what can we do to keep SiteFinderJr from happening?
> >   
> what do you think of as "expensive"?
> 
> I can register a name for a year for the price of 2 beers 
> (ok, beer in Norway is expensive too...)
> 
> There are very few components of setting up something useful 
> on the Internet that cost less.

I don't want to address the current cost, but I will point out that there are 
people who are proposing ideas that would make names cost very much more by 
requiring domain name holders to be authenticated.

I don't like this idea despite the obvious advantages for control of Internet 
crime should it succeed. Running a serious authentication process with global 
scope has significant costs. I don't want to have to pay the cost of an EV SSL 
certificate just to own a domain name. Nor I suspect do most domain name owners.


The core DNS is the one component of the Internet that has to be there for an 
Internet service to function. You can implement multiple redundancy, get drops 
from multiple ISPs etc. but if core DNS is down you are down. People could 
argue that maybe VeriSign has built out for one or two 9s more than is 
absolutely essential, that the Internet only needs to be Internet grade 
reliable not significantly better than carrier grade. That is an interesting 
conversation to have, I don't think that there will be many people in the 
policy community who are receptive to the argument though.
___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Emin Gun Sirer]

> what do you think of as "expensive"?

Anything that has 1000% or higher markup. There is also another kind of
expense: solving the SiteFinder problem took a lot of time, public
outcry and moral outrage from a large group of people. It would have
been nice to just scoot over to a competitor. These are the arguments
for providing competition in a single namespace, as opposed to trying to
solve problems by legislation and public outcry.

Probably won't help with the price of Norwegian beer though.

>Both Milton Friedman and JM Keynes are dead now, ...

Let's not torque the discussion off topic. Free market economics does
not come to bear on the issue because there is no free market to speak
of for registries.

>Or in other words, if IBM wants to keep ibm.com then the root
>must remain under the control of a single exclusive authority.

Agreed. There should be a single logical root, with an entity that has
exclusive authority over it.

>However, to back up a step, what is it that you actually need/want?

A quick question: Right now, we'd like to have a domain delegated to a
large number (say 100+) of nameservers. The registrars we have gone
through impose a limit on the number of nameservers they are willing to
accept. Is this a limit stemming from the .COM registry, or is there a
registrar out there that will let us delegate a .COM domain to a few
hundred nameservers? (And no, not all nameservers need to be returned in
response to every query. A random sampling would be fine). 

Gun.





___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: The IESG Approved the Expansion of the AS Number Registry

[Joe Abley]

On 29-Nov-2006, at 08:30, Henk Uijterwaal wrote:


[EMAIL PROTECTED] wrote:


On the NANOG list it has already been pointed out that a lot
of network management software cannot handle such notation and
in some cases, 1.0 could be interpreted as the IP address 1.0.0.0.  
It has been confirmed that one widely used PERL library interprets  
x.y as IP address x.0.0.y.


I think this is a bug.


If it is, it's a very long-standing one. For example, see INET(3)  
which I think is of 4.2BSD vintage, and which appears to have similar  
semantics to the mentioned perl library:


INTERNET ADDRESSES
 Values specified using the `.' notation take one of the  
following forms:


   a.b.c.d
   a.b.c
   a.b
   a

 When four parts are specified, each is interpreted as a byte of  
data and
 assigned, from left to right, to the four bytes of an Internet  
address.
 Note that when an Internet address is viewed as a 32-bit  
integer quantity
 on the VAX the bytes referred to above appear as ``d.c.b.a''.   
That is,

 VAX bytes are ordered from right to left.

 When a three part address is specified, the last part is  
interpreted as a
 16-bit quantity and placed in the right-most two bytes of the  
network
 address.  This makes the three part address format convenient  
for speci-

 fying Class B network addresses as ``128.net.host''.

 When a two part address is supplied, the last part is  
interpreted as a
 24-bit quantity and placed in the right most three bytes of the  
network
 address.  This makes the two part address format convenient for  
specify-

 ing Class A network addresses as ``net.host''.

 When only one part is given, the value is stored directly in  
the network

 address without any byte rearrangement.


Because of this I think it would be useful for the IETF
to publish a draft defining the notation for AS numbers
so that we can either keep it simple or, if a new notation
is to be used, then publicly state the issues of software which  
needs to be changed. Such a draft should really come

from the WG which extended the AS number in the first place.


There is:

  Canonical Textual Representation of 4-byte AS Numbers
  draft-michaelson-4byte-as-representation-02

describing the format of ASN32 and


The draft above received significant operator criticism.

The consensus I saw on NANOGm, for example, was that there was (a) no  
useful reason to be able to distinguish between a 16-bit AS number  
and a 32-bit AS number less than 65536, (b) no good reason to use  
punctuation to separate the most- and least-significant 16 bits of  
the 32-bit ASN, and (c) every reason to think that the most sensible  
representation was just "bigger decimal numbers".



Joe


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Olaf M. Kolkman]

On 28Nov 2006, at 9:36 PM, Edward Lewis wrote:



path MTU and have to be fragmented.  (By-the-way, why is EDNS/RFC  
2671 not

advancing on the standards track?)


For the same reason almost none of the other DNS RFCs have not  
advanced.


Any volunteers for performing interoperability tests and writing the  
reports that are needed for advancement are kindly requested to send  
a mail to [EMAIL PROTECTED]


--Olaf


---
Olaf M. Kolkman
NLnet Labs
http://www.nlnetlabs.nl/





PGP.sig
Description: This is a digitally signed message part
___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

RE: IM and Presence history

[Janet P Gunn]

"Brian Rosen" <[EMAIL PROTECTED]> wrote on 11/29/2006 08:16:35 AM:

> > However, what this subthread demonstrates is
> > that they were conceptually an incremental change, not a giant,
> > discontinuous, intellectual leap.
> >
> > I thought we all knew that.
> Oh, I agree, we knew that.   There are very, very few discontinuous
> intellectual leaps in our part of the universe.  It's hard to name one
that
> happened in the past decade or two.
>
> Can we name any discontinuous intellectual leaps of late in computer
> networking?  Now that I think about it, forget "of late", have there EVER
> been any?
>

Turing machine? (Computers, but not Networking)

The original Ethernet? (not really "discontinuous", but quite a big leap)

Janet
>
> ___
> Ietf mailing list
> Ietf@ietf.org
> https://www1.ietf.org/mailman/listinfo/ietf


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Patrick Vande Walle]

Brian E Carpenter wrote, On 29/11/2006 10:43:
> your question is linked to whether we treat the namespace as a public
> good to be administered for the greater public good, or as a
> commodity to be treated like coffee beans. And that really isn't
> a question for this technological community.
Depends how you look at it. Technological choices do have an impact on
what the society is able to do (or not). It is up to the society to tell
the engineers what it needs/wants. But I agree there are other fora for
this.

Patrick

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Fwd: The IESG Approved the Expansion of the AS Number Registry

[Henk Uijterwaal]

[EMAIL PROTECTED] wrote:


Until all software understands the bigger numbers, people
will want to continue using the 16-bit ones.


The IESG message talked about numbers from 65536 to some
big number. Here suddenly, we see a reference to some
number of bits.


Meanwhile, to encourage the migration to 4-byte ASNs, the RIRs have


Now there is a reference to some number of bytes. What is going
on here?


I think we're mixing up the number of bits reserved and the decimal
representation.  So far, 16 bits were used, or 2 bytes, the extensions
use 32 bits on the wire.  16 bits can be used for an unsigned decimal
number up to 65536, so it does make some sense to use that instead
of a string of 0's and 1's.


On the NANOG list it has already been pointed out that a lot
of network management software cannot handle such notation and
in some cases, 1.0 could be interpreted as the IP address 
1.0.0.0. It has been confirmed that one widely used PERL 
library interprets x.y as IP address x.0.0.y.


I think this is a bug.


Because of this I think it would be useful for the IETF
to publish a draft defining the notation for AS numbers
so that we can either keep it simple or, if a new notation
is to be used, then publicly state the issues of software 
which needs to be changed. Such a draft should really come
from the WG which extended the AS number in the first 
place.


There is:

  Canonical Textual Representation of 4-byte AS Numbers
  draft-michaelson-4byte-as-representation-02

describing the format of ASN32 and

  RPSL extensions for 32 bit AS Numbers
  draft-uijterwaal-rpsl-4byteas-ext-01.txt

describing what has to be changed in RPSL based tools for ASN32.  For
the latter draft, there is no good place in the IETF right now, but I
do welcome comments.

Henk

--
Henk Uijterwaal   Email: henk.uijterwaal(at)ripe.net
RIPE Network Coordination Centre  http://www.amsterdamned.org/~henk
P.O.Box 10096  Singel 258 Phone: +31.20.5354414
1001 EB Amsterdam  1016 AB Amsterdam  Fax: +31.20.5354445
The NetherlandsThe NetherlandsMobile: +31.6.55861746
--

# Lawyer: "Now sir, I'm sure you are an intelligent and honest man--"
# Witness: "Thank you. If I weren't under oath, I'd return the compliment."


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: IM and Presence history

[Marshall Eubanks]

On Nov 29, 2006, at 8:16 AM, Brian Rosen wrote:


However, what this subthread demonstrates is
that they were conceptually an incremental change, not a giant,
discontinuous, intellectual leap.

I thought we all knew that.

Oh, I agree, we knew that.   There are very, very few discontinuous
intellectual leaps in our part of the universe.  It's hard to name  
one that

happened in the past decade or two.

Can we name any discontinuous intellectual leaps of late in computer
networking?  Now that I think about it, forget "of late", have  
there EVER

been any?


circuit switching -> packet switching

Some people have never recovered.

Regards
Marshall




___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf



___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

RE: IM and Presence history

[Brian Rosen]

> However, what this subthread demonstrates is
> that they were conceptually an incremental change, not a giant,
> discontinuous, intellectual leap.
> 
> I thought we all knew that.
Oh, I agree, we knew that.   There are very, very few discontinuous
intellectual leaps in our part of the universe.  It's hard to name one that
happened in the past decade or two.

Can we name any discontinuous intellectual leaps of late in computer
networking?  Now that I think about it, forget "of late", have there EVER
been any?


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: SMTP compared to IM (Re: DNS Choices: Was: [ietf-dkim] Re: Last Call: 'DomainKeys)

[Eric Burger]

Actually, as I fuzzily recall in the 1986 - 1992-ish period, MCImail had a
large presence for business messaging and CompuServe had a lion's share of
consumer messaging.

Before the flames go on, realize that (1) my memory is fuzzy and (2) the
market was seriously fractured.  The large enterprise market was doing the
Notes thing; the small enterprise market was doing the cc:mail, netware,
etc. thing, and interoperability was something that people gave lip service
to.

What a difference five years made!  By 1996, pretty much everyone
interoperated with Internet Mail.


On 11/26/06 10:35 PM, "John C Klensin" <[EMAIL PROTECTED]> wrote:

> --On Friday, 24 November, 2006 10:30 -0500 Eric Burger
> <[EMAIL PROTECTED]> wrote:
> 
>> Or, the reality that with (at the time) a single dominant
>> network provider made the inter-networking point moot.
> 
> Eric, you are being a little cryptic, perhaps unintentionally.
> What do you mean about a single dominant provider and at what
> time? 
> 
> I would add an observation to Dave's about possibly different
> sets of needs by reminding everyone that considerable IM
> functionality (other than presence) isn't new.  We had
> SEND/SOML/SAML from the beginning of SMTP, even though they had,
> IMO, a very short practical lifespan and, even then, were used
> only in limited communities.  We also we had a couple of flavors
> of the "talk" protocol which were certainly heavily used in some
> places.  "Talk" involved a conversational session while SEND et
> al was closer to what we would call a short message service
> today.  Off the Internet and in the land of BITNET/EARN/etc.,
> there was also an end to end short message protocol and
> mechanism that was extensively used.
> 
> None of these supported a presence mechanism in the sense that
> we understand it today.  As a result, one had to bind a user
> identity to a target host in much the way SMTP does, rather than
> having someone attach to the network at any point and announce
> presence and, implicitly, location.  It is arguably those
> presence and mobility mechanisms and not IM itself that is the
> recent development.  To the degree to which those mechanisms are
> what caused IM to take off, perhaps that reinforces Dave's view
> of different services serving different needs.
> 
>  john
> 
>> On 11/22/06 11:13 AM, "Dave Crocker" <[EMAIL PROTECTED]> wrote:
>> 
>>> 
>>> 
>>> Harald Alvestrand wrote:
> There were no alternatives to SMTP on an IP network until
> Instant Messaging came along.
 
 not since X.400 over X.25 died, no. I thought you were older
 than that
>>> 
>>> And there were all of the individual providers that Michael
>>> cited, such as MCI Mail.
>>> 
>>> 
>> but can be seen in IM, and may likely show up in other
>> forms of communication.  Much of this is simply the nature
>> of software.
> 
> It has nothing to do with software and everything to do
> with architecture. IM networks have less problems because
> all the participants share a relationship with the IM
> service providers.
>>> 
>>> It *is* interesting that the diversity of disconnected email
>>> services was viewed
>>> as a basic problem to solve, whereas most of the Internet
>>> user community does not seem to feel the same pressure to
>>> unify IM.
>>> 
>>> Hmmm.  Maybe IM satisfies a different set of needs than does
>>> email.  So we had better be a bit cautious about trying to
>>> generalize implications between them.
>>> 
>>> 
>>> d/
>> 
>> 
>> ___
>> Ietf mailing list
>> Ietf@ietf.org
>> https://www1.ietf.org/mailman/listinfo/ietf
> 
> 
> 
> 
> 



___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: IM and Presence history

[John C Klensin]

--On Tuesday, 28 November, 2006 22:48 +0100 Eliot Lear
<[EMAIL PROTECTED]> wrote:

> Brian Rosen wrote:
>> If you squint hard enough, everything has already been
>> invented.  Telegraph operators had a form of presence if you
>> squint hard enough.
>> 
>> Presence is a continuously updated 'display' of a set of
>> other people's status.  Finger didn't do that.  Yeah, you
>> COULD have used the mechanism to implement a form of
>> presence, but I don't remember anyone ever doing that, and if
>> they did, it didn't make anyone sit up and take notice like
>> the IM folk's buddy status systems did.
> 
> Mel Pleasant wrote a program for the DEC-20 called "watch",
> which was commonly used on many -20s at the time (this goes
> back to at least the early 80s).  You would provide a list of
> individuals you were interested in watching and the program
> would sit on top of your EXEC and occasionally burp out
> messages that So-And-So has just {logged
> {in|out}|attached|detached}.  At Rutgers we had a program that
> sat on the consoles beneath OPR that would spit out login and
> logout messages of anyone who had wheel.
> 
> Now if you combined Watch with Toggle, a program that let you
> blat a one line message to someone (it also TREPLACEd the
> EXEC) you had many of the same IM features you have today (no
> graphical smileys, bold or italic facing, or direct file
> transfers).

And there were versions of either finger or whois servers
(probably both) that had "continuous" options.  I would still
claim that today's presence models are a significant change,
especially when they are adapted in a distributed
independently-operated server environment and that real-time
messaging is not.  However, what this subthread demonstrates is
that they were conceptually an incremental change, not a giant,
discontinuous, intellectual leap.   

I thought we all knew that.

   john



___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Harald Alvestrand]

Emin Gun Sirer wrote:


As an Internet user, I wonder about two things in the long term:
- why is it so expensive to register a name?
- what can we do to keep SiteFinderJr from happening?
  

what do you think of as "expensive"?

I can register a name for a year for the price of 2 beers (ok, beer in 
Norway is expensive too...)


There are very few components of setting up something useful on the 
Internet that cost less.


 Harald


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: IM and Presence history

[Harald Alvestrand]

Simon Leinen wrote:

Hallam-Baker, Phillip writes:
  

Incidentally, it does need to be [EMAIL PROTECTED] and not
[EMAIL PROTECTED] Google, Yahoo and co need to stop
trying to turn us into serfs by refusing to allow us to own our own
online identity. Stop trying to make a service sticky by making it
costly to switch providers. 

 
I don't know about Yahoo! and co., but Google has a

"bring-your-own-domain" version of some of its services, including
Gmail, see: http://www.google.com/a/ - which allows exactly that.

Disclaimer: This is in beta like so much of Google's stuff, and new
users have to be "approved" - worked instantly for me.  And I don't
have Google shares or something... just a happy user (I almost said
customer... it's easy to forget that Google's users aren't their
customers but their "merchandise" :-).  And my employer provides
DNS-related services.
  
one nice thing about the schema/protocol being part of the naming scheme 
is that it does *not* tie me to a single provider for all services - my 
jabber service for [EMAIL PROTECTED] is provisioned from someone 
who's got no relationship at all to my mail and web services.


Harald


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Michael . Dillon]

> > You can solve the problems in various ways (see Emin Gun Sirer's
> > message) but most of them create a "super-registry" on the top of R1
> > and R2 and you are back to the unique registry model.
> 
> This is a false statement. A basic course on distributed systems will
> cover lots of design alternatives where R1 and R2 are symmetric,
> mutually distrusting and there exists no super-registry, yet there is a
> way to establish whether R1 or R2 acquired the name first. 

This is the problem of using complex and sophisticated 
technical arguments against shared registy models. The fact
is that the domain naming service delegates exclusive control
over a subdomain to a single entity. From that model, companies
like IBM are assured total and exclusive control of the subdomain
ibm.com. If you change that model, then IBM ceases to have
such exclusive control.

I note that numerous organizations have taken an interesting
subdomain and used it to delegate sub-subdomains to other parties.
Such organizations can even share the sub-domain if they wish to.
But doing that infects the entire sub-tree of their domain with
the new model. In order to preserve the possibility that some
domain owners can have total and exclusive ownership we need 
to maintain a single authoritative and exclusive owner of the 
root.

Or in other words, if IBM wants to keep ibm.com then the root
must remain under the control of a single exclusive authority.
 
Fancy technology cannot change this.

--Michael Dillon


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Fwd: The IESG Approved the Expansion of the AS Number Registry

[Michael . Dillon]

> But, with the expanded space, there is an issue of how to transition
> to the larger numbers. This is a software problem as much as
> anything. 

Indeed, there is a software issue here which does not
seem to have been carefully considered.

>Until all software understands the bigger numbers, people
> will want to continue using the 16-bit ones.

The IESG message talked about numbers from 65536 to some
big number. Here suddenly, we see a reference to some
number of bits.

> Meanwhile, to encourage the migration to 4-byte ASNs, the RIRs have

Now there is a reference to some number of bytes. What is going
on here? Is this a question of moving the maximum number
from 65535 to something much larger or is it a matter of
creating new notation to reflect the details of the BGP
protocol change?

Some people have been pushing to make the internal details
of the BGP protocol externally visible even though the new
ASNs are defined in such a way that any 32-bit numbers which
happen to be equal to a 16-bit number are treated as if they
were the old 16-bit number. In other words, if you were allocated
64999 as a 16-bit ASN, you have the right to use 64999 as a
32-bit ASN.

Because of this, some people are demanding that a new notation
be developed to place a punctuation character, either a dot
or a colon, between the two 16-bit segments or between the
2nd and the 3rd byte, if you want to count bytes. Using this
system, there can be no such thing as AS 65536 as was stated
in the IESG message. Instead, that 32 bit quantity will be 
referred to as 1.0 or 1:0.

On the NANOG list it has already been pointed out that a lot
of network management software cannot handle such notation and
in some cases, 1.0 could be interpreted as the IP address 
1.0.0.0. It has been confirmed that one widely used PERL 
library interprets x.y as IP address x.0.0.y.

Because of this I think it would be useful for the IETF
to publish a draft defining the notation for AS numbers
so that we can either keep it simple or, if a new notation
is to be used, then publicly state the issues of software 
which needs to be changed. Such a draft should really come
from the WG which extended the AS number in the first 
place.

--Michael Dillon


___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf

Re: Something better than DNS?

[Brian E Carpenter]

Emin Gun Sirer wrote:

Stephane & Phillip,

I'm thinking of writing a short report that summarizes the invaluable
discussion here and beefing up the system sketch. I think we now agree
that it is possible to have multiple operators manage names in a single,
shared namespace without recourse to a centralized super-registry. 




Do you think it fits well in Hallam-Baker, Phillip's "logical registry"
model?



Yes, the registrars together implement a logical registry without any
centralized component.


I would prefer it if you referred to a logical namespace. Whether or
not there is a registry is a bit of an implementation issue.

...

As an Internet user, I wonder about two things in the long term:
- why is it so expensive to register a name?


Both Milton Friedman and JM Keynes are dead now, but really your
question is linked to whether we treat the namespace as a public
good to be administered for the greater public good, or as a
commodity to be treated like coffee beans. And that really isn't
a question for this technological community.

Brian

___
Ietf mailing list
Ietf@ietf.org
https://www1.ietf.org/mailman/listinfo/ietf