Hi Ran,
On 26th November, you wrote:
> On 24 Nov 2008, at 23:30, Robin Whittle wrote:
> % Ran rejects my critiques but doesn't respond
> % to them with detailed explanations.
>
> All,
>
> Robin's critiques are based on two items.
>
> 1) Robin objects to any host stack changes,
> which is a position held by some portion of the RG.
Yes.
> 2) Robin either hasn't read or doesn't understand
> (I can't tell which) the set of sundry ILNP documents --
> including the published peer-reviewed research papers.
You still haven't provided:
1 - An introduction and analysis of exactly how ILNP would
work, and how it would solve the routing scaling problem
for IPv6. (It can't work for IPv4.)
2 - A complete list (URLs of freely available copies) of all the
documents which are important for understanding ILNP.
I did the best I could to understand your proposal based on the
limited material you have so far made available.
> Detailed responses aren't sensible (respectively) because:
>
> 1) Opinions vary within the RG on whether host stack changes are
> reasonable. That issue is one of opinion, rather than provable
> math/science.
If you want to promote a host-based solution, then it would be good
for you to read and respond in *detail* to my critique of any such
solutions:
Fundamental objections to a host-based scalable routing solution
http://www.irtf.org/pipermail/rrg/2008-November/000233.html
It is not much help to dismiss the opinions of other people on this
matter as somehow less significant because in your opinion, they are
not based on math/science.
I fully support what Noel wrote recently:
http://www.irtf.org/pipermail/rrg/2008-December/000493.html
I think the past decade or so (as far back as ROAD, in some
sense) has taught us that replacing the internetwork layer
with something superior is not easy. Whether or not one
believes that IPv6 is viable on an architectural level (a
discussion I don't want to derail into), its deployment history
has made us all aware that 'forklift upgrades' don't work, and
that viable deployment strategies for anything new have to have
economic benefit for early adopters.
In fact, I'd now say that when analyzed from a practical angle,
what's _most_ important about any replacement piece of the
internetwork architecture is not how nice a design it is, but
how good its deployment path is - including especially the very
earliest stages. I might have the world's best clean-slate
design, but if I don't have a deployment plan which is
economically feasible, it's just a very mildly interesting
piece of paper.
Deployability can't purely be considered from a hard-edge
math/science perspective. It depends in large part on business
models and human behaviour.
> 2) Robin keeps making incorrect assumptions about ILNP and
> then criticises *something other than ILNP* based on his private
> incorrect assumptions, and then mislabels it as ILNP criticism.
I am criticising my perception of ILNP with the intention that you
respond by correcting any misunderstandings. This is standard procedure.
> Robin,
>
> % why isn't ILNP simply a modified stack, with no
> % requirement for any changes to applications?
>
> In fact, I've repeatedly told you that it IS simply
> a modified stack with no *requirement* for any changes
> to applications or to APIs.
>
> Further, I have repeatedly said that the API enhancement
> ideas are *orthogonal* to any specific proposal before
> the Routing RG, would be equally applicable to any proposal,
> and would be applicable even if no protocol changes happened
> in future.
OK - so I understand that you are proposing ILNP in the form of IPv6
stack changes only (no API or application changes) as a complete
solution to the IPv6 routing scaling problem. I understand that you
believe it is a complete solution to the Internet's routing scaling
problem because you think that all Internet users will migrate to
IPv6 sufficiently soon that the IPv4 routing scaling problem will not
get too bad.
> % It is not right to criticise me for not being able
> % to read your mind, or imagine how you might design something.
>
> I've several times suggested you go read the research
> literature, which suggestions you seem to have ignored.
> If and when you do that, you might gain a more complete
> understanding. In an IRTF RG, participants are expected
> to read the relevant research literature. This is not an
> IETF WG, it is an IRTF RG, so IRTF customs and expectations
> apply.
You haven't pointed me to any of these sources which you expect me to
know about, read and understand.
> % If you want me to consider your ILNP proposal seriously, ...
>
> I've repeatedly said that ILNP is a *research project*,
> that the research predates the current Routing RG focua,
> and that I don't really care whether the Routing RG
> endorses it or not.
I assume that because you are mentioning it in the RRG that you
believe your proposal can either solve the routing scaling problem,
or make a significant contribution to its solution.
> Instead, like many researchers, my goal is simply to try
> to put some ideas in front of folks in the hopes of
> stimulating further ideas in other folks' heads. Other
> folks here, for example JNC, operate in a similar mode.
> I find such notes, from anyone, helpful, whether or not
> I might initially agree with all of the concepts/ideas
> put forth.
Sure, I fully support people discussing ideas which are not yet fully
formed. I don't think you have adequately discussed your proposal yet.
> Fundamentally, your main complaint is that ILNP requires
> host stack changes. The *entire* mailing list understands
> that you object to any host stack changes. Repeating it
> over and over (as you have been doing) really isn't helpful.
The fact that you and I have different views on this matter is of
minor interest. What is of great interest to other folks is why we
believe these things. I did my part in explaining my position.
You still support host changes. You haven't responded to my
critique. I think it would be helpful if you did.
> People already know that is your position. And taking that
> position is fine; some folks agree with you; some disagree;
> others haven't taken a firm stance yet.
>
> If you were going to persuade some folks (or have persuaded
> some folks), then likely that would be done (or has been done)
> by your first note to that effect. The repetition is just
> wasted list bandwidth.
I don't think I have been repetitive. I was hoping to prompt people
such as yourself into debating the issues.
> As I said last time, please feel free to ignore ILNP
> if it improves your happiness or lets you spend your time
> on matters you consider more important.
Until you provide a good set of reasons why host-changes are
acceptable, including by responding to my critique of such changes,
and until you provide me with the complete list of the materials you
consider necessary to understand your proposal, I will gladly accept
your offer.
I guess some other folks feel the same way.
> Oh, and I have talked with Tony about his "process" note.
> He has told me point blank that is was NOT directed either
> at me or my notes to the RRG list.
OK. But I find my attempts to understand and critique your proposal
do not result in you providing detailed information about your
proposal, or about why you think it will be practical to solve the
Internet's routing scaling problem with a complete migration to IPv6
with changes to the host stacks.
Good discussions on a mailing list involve responding to critiques in
detail, not just fobbing them off, or labelling them as uninformed
without detailed discussion to correct the supposed misunderstanding.
- Robin
_______________________________________________
rrg mailing list
[email protected]
https://www.irtf.org/mailman/listinfo/rrg
