W3C home > Mailing lists > Public > public-xg-webid@w3.org > January 2012

Re: Certificate Triplify Challenge

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Wed, 11 Jan 2012 21:54:10 -0500
Message-ID: <4F0E4B52.30509@openlinksw.com>
To: public-xg-webid@w3.org
On 1/11/12 8:06 PM, Henry Story wrote:
> Please spend some time READING THIS CAREFULLY.
> Wait until tomorrow before answering. Think over it tonight.
> There is a lot in here, it is easy, but it would be a pitty if because 
> of attempting to read it too quickly the point ended up being lost.

I am reading it CAREFULLY. And responding IN KIND.

> On 12 Jan 2012, at 00:27, Kingsley Idehen wrote:
>> On 1/11/12 6:07 PM, Henry Story wrote:
>>> On 11 Jan 2012, at 20:56, Kingsley Idehen wrote:
>>>> On 1/11/12 2:45 PM, Henry Story wrote:
>>>>> That depends on what your definition of the Web is. Tim Berners 
>>>>> Lee's definition was "a mapping from URIs onto meaning"
>>>>> http://blogs.oracle.com/bblfish/entry/possible_worlds_and_the_web
>>>>>> > >  The end destination is inevitable. WebID or NetID, note, 
>>>>>> I've seen this movie before.
>>>>>> > >  I say "Check!" so your move next:-)
>>>>> Well if the game is a good spec, then my answer is above. (but 
>>>>> perhaps let's not get sidetracked in the meaning of the web, part).
>>>> If you want to hone into definitions, then start with URIs.
>>>> URI abstraction enables one associate URIs with Descriptor 
>>>> Resources for their Referents.
>>> That is not a very clear definition of URIs but I understand what 
>>> you mean.
>>> There are many ways you can put this. URIs are names for things.
>> Yes.
>>> ( Sometimes those things are documents, sometimes not. )
> Notice how I speak of things here, not names.

"URI as Names for Thing"  you said.

Then you say: "( Sometimes those things are documents, sometimes not. )"

>> There lies the problem.
> Which problem? This is by definition how URIs work.

At this point no dispute re. URIs used to name things.

DISPUTE: a Document is a Thing. Any object with discernible 
characteristics is a "Thing" to the observer. Formally, they are 
"Observation Subjects" to the collector of data.  This is a fundamental 
activity in science too re. data collection as part of the observation 

The Problem:
You are trying to claim that a Document and a Thing are Distinct.

What you are really trying to say, with regards to URIs is that a URI 
can be used to Identify a Thing, in a generic sense. It can also be used 
to name a Thing that combines Characteristics and Function (or Role). In 
this case an Address is a Thing, but it has a very specific purpose 
(real world or cyberspace).

>> Sometimes they are Names and sometimes they are Addresses.
> I was speaking about the object, and you have come back round to speak 
> about the name. That is what is confusing you.

No, you remain confused. I believe you are confused because you think 
"object" means object of an S-P-O triple. No, that isn't the sense of my 
> BBLFISH:  "The object to which the name refers is sometimes a document 
> sometimes not."
> KINGSLEY: "The name is sometimes a name sometimes and address"

No, here is what I mean. A URI is a Name in the generic sense. It 
"identifies" something.
A URL is a URI that has a specific purpose/function i.e., it Identifies 
a Location.
A URL is a subClassOf URI. Thus, it is still an "Identifier".

HTTP 200 OK is simply how an HTTP Server acknowledges that a URI is an 
Address (URL). Basically, that the Location it Identifies has something 
to transmit !

> We are not speaking of the same thing. The subject of those two 
> sentences is different.

Exactly! And that isn't my sense. You've basically confirmed the SPO 
triple sense. Not my point.

> In my case the subject is the "object", in your case the subject is 
> the name.

No, in my sense the Subject is the referent of an Identifier. Put 
differently, the Subject is Named by an Identifier. The Subject doesn't 
exist as a physical artifact in the WWW, for instance.

The "S" in an SPO triple has a Referent. That Referent isn't the Object 
of the triple. I don't mean that, and I don't think you mean that either.

> So you shift to the URI, and speak about how it is sometimes a Name 
> and sometimes an address - i.e.: you are splicing the URIs up into 
> URLs and URNs. Where I was talking about the World you are talking 
> about the Word. (notice already how close those two words are:
>     WORLD
>     WORD

No, but we are getting close. I am saying a URI is abstract, but there 
are purpose specific subClasses:

1. URN - Name
2. URL - Address.

When I talk about Name/Address ambiguity, I am talking about the problem 
uniquely introduced by the Linked Data pattern where an HTTP URI (a URL 
to most, based on the dominant WWW usage pattern) serves the role 
naturally handled by a URN. Basically, a de-refrencable URN -- but that 
hasn't materialized due to Browser incomprehension of URI abstraction.

> There is just one L missing between them. The L for Language? Words 
> are about objects in the World. (english is a funny language like that)

You are speaking about English. I am speaking about Computer Science. 
Not exactly the same thing. Please re-read the post from TimBL re. the 
origins of "R" in URI [1]. Also read about the WWW as a distributed 
object systems [2]. That's consistent with what I am talking about.

> At the level of syntax, of language, of tokenised strings, which is 
> where URIs exist, they are all identifiers. "URI" after all is an 
> abbreviation for "Universal Resource Identifiers". Historically there 
> was a confusion before the URI spec came out between locators and 
> names, which led to endless conversations such as this one. So the URI 
> spec was developed to overcome that, which is good.
> URIs then REFER to things such as documents or things that are not 
> documents.

Sorry, but that sentence doesn't really make any kind of sense! A 
"Document" is a Thing too!

> That is the role of URIs to identify things. Those things are (in most 
> cases) not themselves URIs.

Again, you really like to misrepresent. Don't misrepresent my 
statements. Have I told you that a URI doesn't Identify things? How have 
I inferred that in any sense?

An Address is a Name to. Put differently, an Address is an Identifier 
too. A URL is an Identifier too. Just as a Human is a Mammal. Why? 
Because a URL is a subClassOf a URI.

> So let us add a couple of illustrations to help out.

I don't need an illustration from you. I didn't learn what Identifiers 
mean via the WWW or the Semantic Web or Linked Data. If Identifiers, 
unary operators for de-reference and address-of did not exist, there 
wouldn't be a WWW and we wouldn't be exchanging email right now.

The only innovation here is the use of URIs to facilitate the 
functionality of pointers at network scale. That's it!

> In the grey box below we have put the URI which is the name for a 
> document served by an apache server. The name is one thing (it has 21 
> characters) the document is another thing (it contains a public key 
> and many more characters, and can change over time)
> When, as in this case, the URI _refers_ to a document, the sense and 
> the reference of the URI coincide: they are both the document, or 
> information resource referred to.
> But with "https://bblfish.net/#hjs" things are different. That URI 
> refers to <https://bblfish.net/#hjs> ie to me. The sense of the URI 
> can be found at <https://bblfish.net/> which is a document, the same 
> document we were discussing in the previous diagram.

In computer science terms e.g., 'C':

1. https://bblfish.net/#hjs -- "*" (de-reference / indirection unary 
2. https://bblfish.net/ -- "&" (address-of unary operator).

In Web parlance:

1. https://bblfish.net/#hjs -- HTTP URI based Name
2. https://bblfish.net/ -- HTTP URI based Address or Locator (a URL).

Via indirection:
https://bblfish.net/#hjs  gets you to: https://bblfish.net/ 

By introspection style reflection:
https://bblfish.net/ <https://bblfish.net/#hjs> unveils 
https://bblfish.net/ <https://bblfish.net/#hjs>#this .

> Still notice that there is no ambiguity.
> "http://bblfish.net/" refersTo the document.
> "http://bblfish.net/#hjs" refersTo [ foaf:name "Henry Story" ] .

The ambiguity has never been there for me!

> "http://bblfish.net/" sense <http://bblfish.net> .
> "http://bblfish.net/#hjs" sense [ #hjs asgivenby <http://bblfish.net> ] .
> Good so your classification of URIs into names or locators is one that 
> is very prone to lead you and your customers to  confusion if you are 
> not very careful, and this for a number of reasons:
>  "https://bblfish.net/#hjs" is a URI, but is it in your terminology a 
> Name or an Address?


> Does it matter?


> It refers to me and I do have an address - I live in France.

It identifies the location of a resource that bears the representation 
of its referent. Said representation is retrieved via an HTTP GET.

I can also GET the same data via: https://bblfish.net/ 
<https://bblfish.net/#hjs>, that happens every nano second on the WWW of 
Linked Documents.

Linked Data is about additional fidelity, so you need to know the 
Referent. In doing so, you have the ability to exploit equivalence 
fidelity. This is why OWL delivers properties such as owl:sameAs, 
inverseFunctionalProperty etc..

> Ah but then you will say that is not the address you mean: you mean an 
> internet address like "https://google.com/" right? (Ok, but you see 
> how you will already have confused some of your clients here who 
> intuitively will think of "address" to mean the place in physical 
> space) So by address you mean location in the Web Information space. 
> (which is not a space, because everything is one click away - it would 
> be as if every object in physical space were just one step way - true 
> in black holes perhaps but not elsewhere usually).

> But then you get into a second dilemma, because you say ahha 
> "https://bblfish.net/" could also not be an address.

I never said a slash URI couldn't be an Address. You are confusing 
yourself and the matter at hand.

I said: a slash based HTTP URI could be used a either a Name oriented 
Identifier or a Locator/Address oriented Identifier. And modulo 303 you 
end up with Name/Address ambiguity re., this style or HTTP URI !

> It could be a referent to Henry if there is a 303 redirect!!

Tell me about it. And you asked me to READ? You make reference to "Black 
Holes" etc.. In what dimension do you dwell?

> Well this just shows that your address/name distinction is not a 
> syntactic one, but a concept that shifts between syntax and semantics.


> But the reason syntax and semantics are kept stricly separate is to 
> avoid those types of issues, which lead to the confusions that keep 
> dogging us. And I know you have a lot better work to do than keep 
> having everybody you talk to fall into these confusions, and so do 
> most people in the semantic web.

As I already told you. I am not going to get you there. I leave this to 
Jurgen. He understands what I am say. Thus, why not complete the 
learning process via him. You have a lot to LEARN.

Remember, LIFE is LEARNING, so there's nothing wrong discovering that 
you need to LEARN.

> These are really quite easy to understand, but you have to be firm in 
> maintaining them. It is a bit like slavery: one just has to say no, 
> this is a bad concept, it creates confusion between human beings and 
> property, and so we should not maintain that distinction. Simple 
> shifts like that make for revolutions.

Yes, and again, therein lies your problem, you are thinking about a 
REVOLUTION, I am thinking about an EVOLUTION from one realm of computer 
science application to another, courtesy of WWW + Internet ubiquity. 
It's still good old computer science. This you will find out inevitably.


1. http://lists.w3.org/Archives/Public/www-tag/2009Aug/0000.html -- how 
the "R" in URI came to be
2. http://www.w3.org/People/Connolly/9703-web-apps-essay.html -- WWW & 
distributed objects (old post by @DanC)
3. https://plus.google.com/112399767740508618350/posts/U4u4FBmLnvx -- 
understanding distributed data objects.

> Hope this helps,

Yes, re., shedding light on your confusion :-)

> Henry
> So to follow one perhaps we can move to the question you asked in this 
> e-mail 
> http://lists.w3.org/Archives/Public/public-xg-webid/2012Jan/0515.html
> [[
> What's wrong with the following in SAN?
> URI=
> http://id.myopenlink.net/dataspace/person/KingsleyUyiIdehen
> RFC822
> Name=kidehen@openlinksw.com
> ]]
> I answered that here:
> http://lists.w3.org/Archives/Public/public-xg-webid/2012Jan/0532.html
> Given what I have said here, perhaps we can revisit my answer in that 
> e-mail now?
> Henry



Kingsley Idehen	
Founder&  CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen

Received on Thursday, 12 January 2012 02:54:50 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:39:54 UTC