Re: [art] URNs and Last Call: <draft-nottingham-rfc7320bis-02.txt> (URI Design and Ownership) to Best Current Practice

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Jan 7, 2020 at 12:48 PM Barry Leiba <barryleiba@xxxxxxxxxxxx> wrote:  
For the other part of the issue, I'm not fully in agreement with
Phillip and Michael, in that I do think there's a useful distinction
between a URI that names something without specifying how to locate
it... and one that is a locator.  There is, for example, a fundamental
difference between, say, "http:" and "ni:" (see RFC 6920).  

That is precisely what I was talking about when I said that ni: is indexical. But that is no longer an essential distinction either.

We can distinguish names that are under IANA authority (DNS, IP) from those that are not. But that is the limit of it.


Given that it was announced today that SHA-1 was broken, lets take another look at a SIN:


That is a valid email address. It can be routed on the Internet (not the best privacy choice but possible). But it is a name that can be bound to a security policy.

So if  mb2gk-6duf5-ygyyl-jny5e-rwshz is the UDF (SHA-2-512) of Alice's OpenPGP encryption key. We have an implicit semantic here: Encrypt all mail sent to this address with the key that has fingerprint mb2gk-6duf5-ygyyl-jny5e-rwshz

The reason that I have hammered on the fact that the name/locator distinction being bogus is precisely because the obsession with defining a distinction where there is NO difference has caused the neglect of a situation where there is a very real and important one.


A Strong Internet Name allows ANY URI with a DNS component to be turned



On Tue, Jan 7, 2020 at 11:15 AM Michael Richardson <mcr+ietf@xxxxxxxxxxxx> wrote:

Phillip Hallam-Baker <ietf@xxxxxxxxxxxxxxx> wrote:
    > I remain of the view that the URN/URL distinction is unhelpful because it is
    > meaningless. There are no sharp boundaries between categories of identifier
    > as the distinction supposes.

    > Before making this argument in more detail, I will make a plea for respect. I
    > find that the biggest obstacle to getting clarity on naming is that when
    > faced with an analysis that contradicts deeply held beliefs, the usual
    > response is to tell people to 'do some research' or 'understand the issues'.

    > The URN/URI distinction does not in fact represent an essential distinction
    > between types identifiers as claimed. It describes a difference in source of
    > authority at best, not an essential property.

I agree with you.

I think that we might have found new ways in which they are meaningfully
distinct had to a uquitious way to resolve the more abstract URNs.
That didn't happen, and I don't imagine that it is coming in the way envisioned.

HTTPs are names but they are a name that embodies a (fairly) unambiguous means of location.

Having had the misfortune to engage in discussion with acolytes of a prominent opponent of post-modernism, I have been writing a defense of postmodernism today. One of the chief arguments made by the postmodernists is that perfect knowledge is an illusion. And I think that is a profound observation that is essential to understanding why the original concept of URNs and for that matter most registration schemes are doomed to fail.

The Web succeeded because of 404 Not Found. The links are scruffy. And that is fine if you are OK with the idea that knowledge can be scruffy, that it will never be perfect. And that is why the Web came out of CERN because (good) scientists understand that there is no absolute knowledge and that what we are arriving at is a better approximation to the truth, not the truth itself. 

There have been plenty of people making that type of argument since ancient times. But until the mid 80s, most scientists had a very rigid epistemology that was essentially Victorian in its outlook. Scientists were serious people who were finding out profound and universal truths and we had all be very very careful not to make mistakes because that 'would be bad'.

I would encourage people to take a look at what research scientists in their 20s and 30s are saying about their approach to science. They are utterly rejecting that view.

What politicians now call 'Western values' and demand other countries observe is a modern construct that is fifty years (racial tolerance) old at most. In some areas, what is now the fixed consensus is less than a decade old (evolve already). The reason we have the nonsense of .com, .org, .net etc is because DNS was designed by people who had an essentially Victorian view of epistemology in which taxonomy represents essential qualities. (Whether the same is true of the extension TLDs I will leave aside)

The reason I and others received (and continue to receive from some) the patronizing responses we do (do some research you ignorant baboon! you don't understand) is that we have profoundly different and fundamentally incompatible epistemologies and some people regard mine as heresy. Both are capable of self examination. But mine is capable of examining theirs, they can only reject mine. So its like me trying to explain complex numbers to someone who insists that you can't take the square root of a negative number and can't get past that.

The route out of formal methods to PoMo is that to solve the problems we faced in the 1980s in formal methods, it was necessary to apply meta-mathematics and understand the limitations of logic. And once you do that you are in Post Modernism.


So the URN notion that we must carefully decide how to resolve before we defined schemes was wrong. It was and is much more important to have a catalog that is as complete as possible than for the entries to be correct.

Resolution of many types of name is inherently ambiguous because the context varies.. We cannot resolve a barcode on a physical object like a can of beans. But we can ask for information on the can of beans and we can order an instance of the can of beans for delivery. And both forms of 'location' are going to be subjective and depend on the context in which they are attempted.

[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Mhonarc]     [Fedora Users]

  Powered by Linux